Full text of "Sulphuric acid ; Alkalies and hydrochloric acid ; Manufacture of iron ; Manufacture of steel"

See other formats

Google

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project

to make the world's books discoverable online.

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover.

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the

publisher to a library and finally to you.

Usage guidelines

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing tliis resource, we liave taken steps to
prevent abuse by commercial parties, including placing technical restrictions on automated querying.
We also ask that you:

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for
personal, non-commercial purposes.

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the
use of public domain materials for these purposes and may be able to help.

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for in forming people about this project and helping them find
additional materials through Google Book Search. Please do not remove it.

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner
anywhere in the world. Copyright infringement liabili^ can be quite severe.

About Google Book Search

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web

at|http: //books .google .com/I

K.F. WENDT LIBRARY

UW COLLEGE OF ENGR.

215 N. RANDALL AVFf^UE

MADISON Wl 53706

INTERNATIONAL
LIBRARY OF TECHNOLOGY

A SERIES OF TEXTBOOKS FOR PERSONS ENGAGED IN THE ENGINEERING

PROFESSIONS AND TRADES OR FOR THOSE WHO DESIRE

INFORMATION CONCERNING THEM. FULLY ILLUSTRATED

AND CONTAINING NUMEROUS PRACTICAL

EXAMPLES AND THEIR SOLUTIONS

SULPHURIC ACID

ALKALIES AND HYDROCHLORIC ACID

MANUFACTURE OF IRON

MANUFACTURE OF STEEL

SCRANTON :
INTERNATIONAL TEXTBOOK COMPANY

Entered at Stationers' Hall, London.

Sulphuric Acid: Copyright, 1902, by INTERNATIONAL TEXTBOOK COMPANV.
Entered at Stationers' Hall, London.

Alkalies and Hydrochloric Acid : Copyright, 1902, by INTERNATIONAL TEXTBOOK
Company. Entered at Stationers' Hall, London.

Manufacture of Iron : Copyright, 1902, by INTERNATIONAL Textbook Company.
Entered at Stationers' Hall, London.

Manufacture of Steel: Copyright, 1902. by International Textbook Com-
PANY Entered at Stationers' Hall, London.

AU rights reserved.

Press of Eaton & Maih3

NEW YORK

104383

MAR 3,0 1907

The International Library of Technology is the outgrowth
of a large and increasing demand that has arisen for the
Reference Libraries of the International Correspondence
Schools on the part of those who are not students of the
Schools. As the volumes composing this Library are all
printed from the same plates used in printing the Reference
Libraries above mentioned, a few words are necessary
regarding the scope arid purpose of the instruction imparted
to the students of — and the class of students taught by —
these Schools, in order to afford a clear understanding of
their salient and unique features.

The only requirement for admission to any of the courses
offered by the International Correspondence Schools, is that
the applicant shall be able to read the English language and
to write it sufficiently well to make his written answers to
the questions asked him intelligible. Each course is com-
plete in itself, and no text^books are requirjed other than
those prepared by the Schools for the particular course
selected. The students themselves are from every class,
trade, and profession and from every country; they are,
almost without exception, busily engaged in some vocation,
and can spare but little time for study, and that usually
outside of their regular working hours. The information
desired is such as can be immediately applied in practice, so
that the student may be enabled to exchange his present
vocation for a more congenial one, or to rise to a higher level
in the one he now pursues. Furthermore, he wishes to
obtain a good working knowledge of the subjects treated in
the shortest time and in the most direct manner possible.

• • •

iv PREFACE

In meeting these requirements, we have produced a set of
books that in many respects, and particularly in the general
plan followed, are absolutely unique. In the majority of
subjects treated the knowledge of mathematics required is
limited to the simplest principles of arithmetic and mensu-
ration, and in no case is any greater knowledge of mathe-
matics needed than the simplest elementary principles of
algebra, geometry, and trigonometry, with a thorough,
practical acquaintance with the use of the logarithmic table.
To effect this result, derivations of rules and formulas are
omitted, but thorough and complete instructions are given
regarding how, when, and under what circumstances any
particular rule, formula, or process should be applied; and
whenever possible one or more examples, such as would be
likely to arise in actual practice — together with their solu-
tions— are given to illustrate and explain its application.

In preparing these textbooks, it has been our constant
endeavor to view the matter from the student's standpoint,
and to try and anticipate everything that would cause him
trouble. The utmost pains have been taken to avoid and
correct any and all ambiguous expressions — both those due
to faulty rhetoric and those due to insufficiency of statement
or explanation. As the best way to make a statement,
explanation, or description clear is to give a picture or a
diagram in connection with it, illustrations have been used
almost without limit. The illustrations have in all cases
been adapted to the requirements of the text, and projec-
tions and sections or outline, partially shaded, or full-shaded
perspectives have been used, according to which will best
produce the desired results. Half-tones have been used
rather sparingly, except in those cases where the general
effect is desired rather than the actual details.

It is obvious that books prepared along the lines men-
tioned must not only be clear and concise beyond anything
heretofore attempted, but they must also possess unequaled
value for reference purposes. They not only give the maxi-
mum of information in a minimum space, but this infor-
mation is so ingeniously arranged and correlated, and the

PREFACE V

indexes are so full and complete, that it can at once be
made available to the reader. The numerous examples and
explanatory remarks, together with the absence of long
demonstrations and abstruse mathematical calculations, are
of great assistance in helping one to select the proper
formula, method, or process and in teaching him how and
when it should be used.

Three of the volumes of this library are devoted to sub-
jects pertaining to Applied Chemistry. The present volume
contains descriptions of the following industries : manufac-
ture of sulphuric acid, manufacture of alkalies and hydro-
chloric acid, manufacture of iron, and manufacture of steel.
The manufacture of sulphuric acid, a comparatively new
industry in this country, is increasing rapidly ; the subject
is thoroughly treated and liberally illustrated with detail
plans of the latest constructions and improvements. Manu-
facture of Alkalies and Hydrochloric Acid treats on the
manufacture of sodium chloride, soda, ammonia recovery,
cryolite soda process, sodium sulphate, sodium thiosulphate,
sodium hydrate, hydrochloric acid, chlorine, bleaching
powder, etc., including the latest electrolytical processes
and a description of the analytical methods of intermediate
and finished products. Manufacture of Iron presents
a complete description of modern blast-furnace practice,
and contains besides valuable formulas for the calculation
of blast-furnace burdens. Manufacture of Steel is a
complete review of the art of steel making as practiced in
this country. It has been our endeavor to expound the
dominant principles that govern these industries and give
at the same time a detailed account of the various manu-
facturing processes, with special consideration of the most
modern American practice.

The method of numbering the pages, cuts, articles, etc.
is such that each subject or part, when the subject is divided
into two or more parts, is complete in itself; hence, in order
to make the index intelligible, it was necessary to give each
subject or part a number. This number is placed at the top
of each page, on the headline, opposite the page number;

VI PREFACE

and to distinguish it from the page number it is preceded
by the printer's section mark (§). Consequently, a refer-
ence such as § 30, page 26, will be readily found by looking
along the inside edges of the headlines until § 30 is found,
and then through § 30 until page 26 is found.

International Textbook Company.

CONTENTS

Sulphuric Acid Section Page

Introduction 27 1

Principles Governing the Manufacture

of Sulphuric Acid 27 7

The Production of Sulphur Dioxide or

Burner Gas 27 21

Furnaces and Burners for the Produc-
tion of Burner Gas 27 23

Brimstone Burners 27 25

Pyrites Burners 27 27

Testing the Burner Gas 27 32

Calculation of Volume of Burner Gas . 27 40

The Catalytic, or Contact, Process . . 27 43

The Chamber Process 28 1

Apparatus Employed in the Chamber

Process 28 6

Surface Condensers 28 16

Operation of the Chamber Process . . 28 28
Purification of Chamber Acid .... 28 42
Concentration of Dilute Acid Solutions
and the Production of Sulphuric Mono-
hydrate 28 48

Alkalies and Hydrochloric Acid

Sodium Chloride 29 1

Sodium Carbonate ........ 29 9

The Solvay Process 29 11

vi CONTENTS

Alkalies and Hydrochloric Acid — Cont'd Section Page

Cryolite Soda Process 29 32

Salt Cake .29 34

Soda by the Le Blanc Process .... 29 46

Sodium Hydrate 29 74

Sodium Bicarbonate 29 84

Hydrochloric Acid 30 1

Chlorine 30 8

Bleaching Powder ........ 30 29

Potassium Chlorate 30 37

Electrolytic Methods 30 44

Electrolytic Preparation of Alkali and

Chlorine 30 . 66

Fused Electrolyte . 30 69

Dissolved Electrolyte 30 73

Potassium Chlorate 30 84

Analytical Methods 31 1

Ammonia Soda 31 1

Salt-Cake Process ........ 31 17

Le Blanc Process 31 20

Chance-Claus Sulphur Recovery ... 31 29

Sodium Bicarbonate 31 32

Hydrochloric Acid 31 36

Chlorine, Bleaching Compounds, Chlo-
rates 31 41

Manufacture op Iron

Introductory 32 1

Iron Ores 32 3

Classification of Iron Ores 32 3

Distribution of Iron Ores in the United

States 32 5

Valuation of Iron Ore 32 7

Preparation of Ores 32 7

Fuel for Blast Furnace 32 14

Fluxes 32 16

Blowing Engines 32 18

Stoves 32 19

CONTENTS vii

Manufacture op Iron — Contimted Section Page

Pipe Stoves 32 19

Regenerative Stoves 32 21

The Furnace 32 26

Casting . . / 32 39

Slags 32 45

Calculation of Burdens 32 48

Classification of Iron 32 53

Practical Suggestions 32 61

Manufacture op Steel

Introductory 33 1

The Open-Hearth Process 33 5

Acid and Basic Open-Hearth Systems . 33 23
Gaseous Fuel Used in Open- Hearth

Furnaces 33 30

Natural Gas 33 30

Artificial Gas 33 32

The Acid Open-Hearth Process ... 33 44

The Basic Open-Hearth Process ... 33 60

The Bessemer Process 34 1

The Acid Bessemer Process 34 13

The Basic Bessemer Process .... 34 20

Recarbonization 34 33

The Crucible Process 34 44

Alloy Steels 34 56

Steel Castings 35 1

Defects in Steel 35 6

Effects of the Usual Elements Present

in Steel 35 11

Examination of the Finished Product . 35 15

Recent Progress in Steel Making ... 35 25

Treatment of the Ingot 35 32

Refractory Materials 35 37

SULPHURIC ACID

(PART 1)

INTRODUCTION

1. General Remarks and Definitions. — Before consid-
ering the technology of sulphuric acid, it is of the greatest
possible importance to have a clear idea as to just what sul-
phuric acid is and the place it occupies among the oxides
and acids of sulphur. The technical processes to be
described, instead of seeming complicated will then appear
consequent and logical, and the bewildering chemical and
commercial terminology with which the evolution of the
manufacture has incrusted the subject will be cleared away,
or at least will be more readily understood.

3. Hydrates and Solutions of Sulphur Trioxide. — It

was stated in Inorganic Chemistry that sulphur trioxide SO^
when absolutely pure is a colorless, mobile liquid of 1.940 sp.
gr. at 16° C, and when cooled it solidifies into long, trans-
parent prismatic crystals. If a little water is added, a
mass of opaque, white, asbestos-like crystals will result,
which melt at about 60° C.

If 10.11 per cent, of water is added to the pure sulphur
trioxide, a transparent crystalline mass is obtained, melting
at 35° C. and readily decomposing at moderate heat into
H^SO^ and SO,.

If 18.37 per cent, of water is added to pure sulphur triox-
ide, a limpid, colorless, oily fluid is obtained of 1.8872 sp. gr.

§27

For notice of copyright, see page immediately following the title page.

2 SULPHURIC ACID § 27

at 15° C. (Lunge 1.8385), which solidifies at 0° C. into large,
plate-shaped crystals and readily decomposing at moderate
heat into Hfi and 5(9,.

If 31.04 per cent, of water is added to the pure sulphur
trioxide, large, clear, hexagonal, columnar crystals that
melt at 8.5° C. are obtained.

All these mixtures of pure sulphur trioxide and water, or
solutions of sulphur trioxide in water, possess characteris-
tics, such as crystallization, melting points, change of vol-
ume, etc., that show them to be definite chemical compounds
or hydrates of sulphur trioxide.

Again, if from 14 to 18 per cent, of water is added to pure
sulphur trioxide, a thick, oily liquid that throws off dense
white fumes on exposure to the air is obtained. These
fumes are the vapor of sulphur trioxide combining with the
moisture of the air and forming a non-volatile hydrate.

If 23.67 per cent, of water is added to the pure sulphur
trioxide, a thick, oily liquid is obtained of 1.835 sp. gr. and
stable at ordinary temperatures. This is the oil of vitriol
of commerce, or 66° Baume sulphuric acid (in the United
States).

In the same way, water may be added in other percent-
ages; in some cases hydrates, but nearly always simply solu-
tions, result.

3. If these hydrates exist at low temperatures as definite
crystalline compounds, and if on a rise of temperature they
all decompose with more or less ease with the disengage-
ment of either sulphur trioxide or water, and if in their ordi-
nary form they present all the properties of simple solutions,
it follows that between sulphur trioxide 5(7, and water Hfi
there exists a consecutive series of homogeneous liquids or
solutions, aniong which must be distinguished definite com-
pounds, or hydrates; therefore, it is quite justifiable to look
for other definite compounds between sulphur trioxide and
water, which are distinguished by the variation of proper-
ties of any kind uniformly occurring with a solution of any
uniform percentage of sulphur trioxide and water. Few of

§ 27 SULPHURIC ACID 3

these variations of properties of definite solutions have been
determined with sufficient accuracy.

In other words, the term sulphuric acid is the generic
name of a series of solutions of sulphur trioxide in water,
some of which are chemical hydrates of the sulphur trioxide
and most of which are merely solutions of convenient
strength for use in the arts.

4, In Table I are given the principal characteristics
of the various commercial solutions of sulphur trioxide in
water. The best known hydrates are also shown. It will
be noticed that none of the hydrates are recognized com-
mercially.

5* Nomenclature of Solutions and Hydrates of Sul-
phur Trioxide. — The term sulphuric acid is usually
applied to the monohydrate of sulphur trioxide SO^.H^O^
and yet at the same time it covers the whole range of
hydrates and solutions containing a smaller percentage of
SO^ than the monohydrate, and also the hydrates and solu-
tions containing more SO^ than the monohydrate. As the
moment that moisture is added to sulphur trioxide it
becomes an acid, the term sulphuric acid therefore applies
to the whole range of hydrates and solutions, of SO^ in
water. There is no reason why the monohydrate should
monopolize the term sulphuric acid other than the fact that
it marks the margin of the acids of sulphuric trioxide that
are stable in liquid form at ordinary temperatures ; and even
this is not quite correct, as the actual monohydrate itself,
even at 40° C, begins to give off fumes of sulphur trioxide,
and even in a dry atmosphere becomes weaker until it con-
tains 1.5 per cent, of water. At this point, however, it
really becomes stable, so far as the separation of the sulphur
trioxide is concerned, and in a dry atmosphere will remain
unchanged.

It is this sulphuric acid that contains not more than
98.5 per cent, of H^SO^^ or 80.41 per cent, of SO^ and 19.59 per
cent, of water, that it has been possible to make by the

4 SULPHURIC ACID § 27

so-called chamber process^ aided by concentration (evapo-
ration of water) and by distillation, and which has therefore
been cQmmercially available. If stronger ^cid were required,
recourse to the fortification of this acid by sulphur trioxide
made at great cost was necessary. The 80.41-per-cent. 5(7„
or 98.5-per-cent. H^SO^^ or as near to it as possible, was
fortified with sulphur trioxide until it became 81.63-per-
cent. SO^ acid (monohydrate), and if a greater strength or
a so-called fuming acid were required, more sulphur trioxide
was added, and the acid thus fortified considered as the mono-
hydrate plus a certain percentage of free sulphur trioxide.

6. Nordliausen or Fuming Sulphuric Acid. — As until
comparatively recently the only commercial sulphur trioxide
was produced as a faming: or IN'ordhausen acid (i.e., an acid
containing a greater percentage of sulphur trioxide than the
monohydrate) and very costly to make, every effort was
made to bring the chamber acid to its greatest strength (to
eliminate by evaporation as much water as possible). For,
as the proportion of sulphur trioxide to water in monohy-
drate is 81.63 to 18.37, every part of water in the acid to be
fortified first requires 4.444 parts of sulphur trioxide to form
the monohydrate before any so-called free sulphur trioxide
or H^SO^ -f SO^ is obtained. With the one exception of
pyrosulphuric acid, disulphuric acid, or solid oleum, terms
applied to the hydrate //,5,6?„ or 250, + Hfi^ there is no
nomenclature that covers the whole range of acids from the
monohydrate, or 81.63-per-cent. sulphur trioxide, to the sul-
phur trioxide itself, except the terms fuming or Nordhausen
acids; the first is descriptive of a characteristic of these
acids and the second is the name of a town in Prussian
Saxony where a warehouse for the storage of these acids
was located, the factories being at Braunlage, Goslar, and
other places.

As, therefore, the term sulphuric acid is used not only
to define the actual sulphuric monohydrate, but also to
describe the whole range of hydrates and solutions of sul
phur trioxide, it becomes necessary for accurate expression

§ 27 SULPHURIC ACID 5

to define the hydrate or solution referred to in terms of per-
centage of sulphur trioxide contained in it. When acids
stronger than commercial oil of vitriol (76.33 per cent, of
SO^) were rare and acidum sulphuricum distillatum (80.41
per cent, of SO^) was the strongest. commercial acid known,
it was, of course, natural that the strength of all acids
should be referred to the monohydrate, or nearest, hydrate.

?• Commercial Methods for Determining: the
8tren^h of Solutions Weaker than the Monohydrate.

For ascertaining the strength of those solutions weaker than
the monohydrate, recourse is had to their specific gravity — a
fairly accurate method up to a certain point, but uncertain
just about the reference point (monohydrate), as in passing
from 79.99 per cent, of SO^ (98 per cent, of H^SO^) to 81.63
per cent, of SO^ (100 per cent, of H^SO^ the specific grav-
ity decreases from 1.8415 to 1.8372. The specific gravity,
however, rises just so soon as the monohydrate point is
passed and SO^ is slightly in excess.

In commercial acids a further cause of inaccuracy exists,
owing to the effect on the specific gravity of the almost con-
stant impurities present. Furthermore, commercial methods
of observing the specific gravity are neither uniform nor
accurate, even apart from the inaccuracy of the instruments
themselves.

8. Speeiflc-Gravlty, or Density, I>eterminations. — ^

The hydrometer used in connection with sulphuric acid is
simply an instrument for determining its specific gravity, or
density, in comparison withdistilled water at 15° C. (or 60° F.
in the United States). With commercial acids the use of
the hydrometer should be limited to the solutions contain-
ing up to 76.33 per cent, of SO^ (93.5 per cent, of H^SO^).
Specific-gravity determinations beyond this point are unre-
liable on account of impurities in the acid, and all deter-
minations above this point should be made alkalimetrically.
Apparently, it should be easy to make the hydrometric
scale an exact basis of universal calculation, but in practice
there are many different hydrometer scales. One of the

SULPHURIC ACID

§27

difficulties is the uncertainty as to the standard of maxi-
mum density. In Europe this is generally understood to
be 1.842 sp. gr. at 15° C, or 66° Baum6. As this specific
gravity would correspond to a fuming acid, it is difficult to
see on what this standard is based. The specific gravities
of solutions of sulphur trioxide, just between 97 and 100 per
cent, of H^SO^ (79.19 and 81.63 per cent, of 5(7,), are given

in Table IL

TABIiB n

SPECIFIC GRAVITY OF SOLUTIONS OF SULPHUR TRIOXIDE

NtSOi

SOt

Specific Gravity

97.00

79.19

1.8410

97.70

79.76

1.8415

98.20

80.16

1.8410

98.70

80.57

1.8405

99.20

80.98

1.8400

99.45

81.18

1.8395

99.70

81.39

1.8390

99.95

81.59

1.8385

100.00

81.63

1.8372

9, In England, the Twaddell scale starts with a maxi-
mum specific gravity of 1.850, or 170°. Each intermediate
degree represents a difference of .005 in specific gravity.
In the United States, the Baum6 scale is also used, the 66°,
however, corresponding to 93.5 per cent, of I/^SO^, or
76.3265 per cent, of 5(?„ or a specific gravity of 1.835.

The modulus, or formula of division, where d = specific

gravity and n = the number of degrees, for the European

Baume is

144.3

d =

144.3 -«'

and for the United States Baum^ is

d=^

145

145 - ;/•

§ 27 SULPHURIC ACID 7

Throughout this work, the United States Baum6 is used,
as it is the one universally adopted by sulphuric-acid manu-
turers in this country. In addition to these scales, those of
Gerlach and others are used in different parts of Europe
and in different factories in the same country. All of which
tends to show that the only precise and accurate way of
describing the acids of sulphur trioxide is in terms of per-
centage contents of such oxide.

PRINCIPIiES OOTERNING THE MANUFACTURE OF

8UL.PHURIC ACID

10. When sulphur dioxide SO^ and oxygen are brought
together imder certain conditions, they combine to form
sulphur trioxide SO^, This in the presence of water vapor
becomes hydrated, and these hydrates are known as sul-
phuric acid. The conditions under which sulphur dioxide
and oxygen may combine are varied. For the commercial
manufacture of sulphuric acid, this combination is brought
about in two ways.

1. By what is known as contact or catalytic actioji^ the
two gases are brought together in the presence of certain
substances, as finely divided platinum, and other substances
described farther on, that have the peculiar power to cause
them to unite chemically. The dry sulphur trioxide thus
formed is absorbed in the proper amount of water, to give
an acid of the desired strength.

2. The two gases are brought together in the presence of
steam and some of the higher oxides of nitrogen, as, fof
instance, Nfi^, The oxide of nitrogen gives up oxygen to
the sulphur dioxide and forms, in the presence of water
vapor, sulphuric acid. The lower oxide of nitrogen formed
immediately takes up oxygen from the air present and is
regenerated.

The reaction is quite complicated but is continuous. A
small amount of oxide of nitrogen serves to oxidize an

8 SULPHURIC ACID § 27

indefinite amount of sulphur dioxide to the trioxide. This
is the reaction used in the so-called chamber process.

In the discussion of the two processes for the manufac-
ture of sulphuric acid, the above-mentioned reactions will be
quite fully dealt with. Before discussing these, however,
the various sources of sulphur and the preparation of sul-
phur dioxide will be taken up.

11, Raw Materials Used In the Manufacture of Sul-
phuric Acid. — Commercial sulphuric acid is^ derived from
the following raw materials:

1. Brimstone (a) derived , from sedimentary deposits
accompanied by or derived from gypsum, found in Sicily,
Louisiana, etc. ; (^) derived to a limited extent from
volcanic deposits (Solfatara).

2. Recovered sulphur (a) from alkali waste (Chance and
Klaus processes) ; (^) from spent oxides from gas works.

. 3. Sulphureted hydrogen obtained as a by-product in
the manufacture of ammonium sulphate, etc.

4. Iron pyrites, in which the principal value is the
sulphur.

5. Iron pyrites with copper pyrites, in which the princi-
pal value is copper (sometimes also gold and silver) and the
sulphur may be considered as a metallurgical by-product. •

6. Zinc blende, in which the principal value is zinc.

7. Copper-nickel pyrrhotites, in which the principal
value is the metal.

8. Copperas slate (Vitriolschiefer), which is oxidized to
ferrous and then to ferric sulphate in the Nordhausen proc-
ess for the manufacture of fuming sulphuric acid; also
other acid sulphates of the alkalies, which, upon being
heated, are first changed into pyrosulphates and then split
up into neutral sulphates and sulphur trioxide.

It will be noted that these raw materials divide themselves
into the following classes: {a) Where the sulphur is the
principal or only value, as brimstone and most iron pyrites;
[p) where the sulphur is a recovered or a by-product from a

§ 27 SULPHURIC ACID 9

previous chemical process, and, therefore, only available
locally or under special conditions, as hydrogen sulphide,
alkali waste, etc; (c) where the sulphur is of secondary
value and is virtually a waste product in a metallurgical
operation; {d) where the sulphur is derived from sources
that are only suited on account of their cost for special proc-
esses and products, as the various sulphates.

12. The history of the manufacture of sulphuric acid
commercially shows, as may be expected, that at first brim-
stone, as being technically the simplest raw material, was
exclusively used. This was, in turn, supplanted by iron
pyrites. Iron pyrites are now being largely driven out by
the waste gas produced in the desulphurization of copper,
zinc, nickel, gold, and silver ores, and it is not difficult to see
that in time the great bulk of acid will be produced as an
adjunct to the various metallurgical processes. Literally,
in the United States thousands of tons of sulphur are being
delivered into the air as sulphurous gas every day of the
year by the various metallurgical works. The capital
invested in the present plants, the capital cost of making
the necessary changes to render the gas available, remoteness
from present markets, and other necessary costly adjust-
ments alone prevent this sulphur from being recovered as
sulphuric acid.

As to the use of sulphates for the manufacture of fuming
acid, this industry is practically dead, having been replaced
entirely by the catalytic or contact process described far-
ther on.

13. Preparation of the Raw Material. — Brimstone
or sulphur requires little or no preparation, as it comes to
the market in suitable condition to be put into the burners.
Crude sulphur in the Sicilian warehouses is graded accord-
ing to its purity and also, in a way, according to the method
employed in its extraction.

Grading is done by simple inspection, without sampling or
assaying. Three qualities are recognized : firsts^ seconds^
and thirds. Light-colored sulphurs arfe included in the first

10 SULPHURIC ACID g 27

two grades and darker varieties in the thirds. Seconds and
thirds are subdivided into * * vantaggiata, " * * buona, " and * *cor-
rente.** Firsts are nearly chemically pure and of a canary-
yellow color, while seconds vantaggiata are but slightly
inferior. Seconds buona have a fine chrome-yellow color;
seconds corrente have a dirty yellow color; and thirds are
chocolate brown on the exterior, shading to greenish brown
inside.

14. For the American trade, two ^special classes are
made, seconda uso Aiuerica, best seconds, which is a mixture
of seconds corrente and thirds vantaggiata; and terza uso
Avterica^ best thirds, a mixture of terza vantaggiata and
terza buona. The chemical purity of these classes differs
comparatively little. The various grades of seconds range
from 99.85 to 99.70 sulphur; and of thirds, from 99.64 to
99.58 sulphur. The principal difference — namely, that of
color — is due to temperature and other points connected with
the fusion.

15. The spent oxides of gas works, which contain sulphur,
are first treated for the recovery of the salts of ammonia,
ferrocyanides, and sulphocyanides, and are then roasted as
if they were the fines, or dust, of the metallic sulphides
and in the same class of furnaces.

16. Sulphtireted liydrogren, when ignited in the air,
burns with a blue flame, water and sulphur dioxide resulting

with limited air access, or when the flame is cooled by the
introduction of a cold body; only hydrogen burns and the
free sulphur separates. Advantage is taken of this reaction
to use the hydrogen sulphide produced in the Chance
process for the utilization of alkali waste for the manufac-
ture of sulphuric acid ; or by the Klaus process, for the
recovery of sulphur. The hydrogen-sulphide gas is simply
burned in a suitable combustion chamber and the resulting
SO^ passed to the lead chambers, or otherwise oxidized
to SO^.

§ 27 SULPHURIC ACID 11

17. The metallic sulphides, the bisulphides of iron,
or iron and copper pyrites, can be roasted both in the form
of small lumps or as dust, or fines, and by their own heat
of combustion alone. The monosulphides, or copper- or
nickel-bearing pyrrhotites and zinc blendes must be roasted
as fines and with the aid of additional fuel. Many pyrites
are so friable as to crumble to fines when being mined, and
many pyrites carrying copper, gold, silver, and other valuable
metals are in the form of concentrates, or fines; such ores
are disseminated, when found, among large proportions of
quartz or other gangue matter, or consist of the sulphides
of several metals, which it is desirable to separate before
further metallurgical treatment.

If these ores occur in massive form, they must first be
broken into small pieces. This is done either by hand or by
rock breakers. The method used will depend on local condi-
tions, such as cost of labor, etc., and on the mechanical
condition of the ore, such as friability, etc. The ore must
then be screened and sized. As a rule, except in the case of
a very free-burning iron pyrite or under special conditions,
such as extreme friability of the ore and insufficient
facilities for roasting the fines^ the largest size produced
should pass through a 3-inch ring ; the next size should pass
through a 2-inch ring; and so on. Too much emphasis
cannot be given to the necessity for properly sizing the ore
and burning one size only in the same burner. This applies
not only to the lump ore but also to the smalls and fines.

18. In the first place, it is evident that for a **dead"
roast, or a roast of equal efficiency, the capacity of any given
furnace will be controlled by the time taken to roast the
largest pieces. Therefore, to secure the efficient and eco-
nomical use of costly apparatus, the economy of power and
labor, or in other words, maximum output at minimum
cost, it is necessary to have a reasonably close sizing of the
charge of raw ore to any given furnace. Moreover, that
serious class of troubles met with in roasting ores, called
clinkering^ scarrings etc., and much of the labor of breaking

12 SULPHURIC ACID § 27

up and barring the bed of ore in a lump burner is the direct
result of improper sizing. These scars, or clinkers, are really
the formation of a fusible matte of ferrous sulphide FeS^
owing to the irregular passage of air through the bed of ore
on the furnace grates. If the ore is reasonably sized, air
will be uniformly admitted through the bed and each piece
of ore will get sufficient air for its complete oxidation.
Moreover, the resulting regularity in the condition of the
furnaces will tend to produce uniformity in the conditions of
the burner-gas and the acid-making process.

19. Combustion of Sulphur and Its Thermoclieni-
Istry. — When brimstone or a metallic sulphide is heated in
the air, or burned, the following reaction takes place:

5, + 2C^, = %S0^

In this respect, the combustion of sulphur appears to
form an exception to the general rule of thermochemistry —
viz., that where two or more compounds are possible as the
products of chemical combination, that product will be
formed which produces the greatest heat in the reaction ; for
example, C and O can form carbon monoxide CO or carbon
dioxide C(?„ and carbon dioxide is the usual product of com-
plete combustion; sulphur and oxygen can form sulphur
dioxide SO^ and sulphur trioxide 5(?„ yet sulphur diox-
ide is the usual product of combustion. The reason for this
is that the heat of the oxidation of sulphur to the trioxide is
so great as to cause the dissociation of the trioxide into the
dioxide and oxygen, or in other words, that the difference
in the temperature of the production and dissociation of
sulphur trioxide is so slight that unless some means are taken
to carry off the heat of the reaction effectively it cannot
exist. This fact becomes highly important in the considera-
tion of the various contact processes.

As a matter of fact, the gas produced by the combustion
of brimstone or the metallic sulphides always contains
varying proportions of sulphur trioxide, so that techni-
cally the equation given above does not quite represent the
reaction of the combustion of sulphur in air.

§ 27 SULPHURIC ACID 13

The fact that the burner gas contains varying quantities
of sulphur trioxide is shown by the formation of free sul-
phuric acid, when such gases are washed in water or dilute
sulphuric acid or passed over iron filings before being used
in the manufacture of sulphite pulp.

30. Burner Gas, — Burner gas, whether derived from
the combustion of brimstone or the metallic sulphides, forms
the basis of the manufacture of sulphur trioxide and all its
hydrates and solutions. It consists, according to the raw
material used, of a mixture of sulphur dioxide and sulphur
trioxide, nitrogen, oxygen, and many impurities, such as flue
dust, iron, silica, arsenious and hydrofluoric acids, and com-
pounds of selenium, thallium, zinc, lead, etc.

21. As air consists approximately of 79 parts, by volume,
of nitrogen and 21 parts, by volume, of oxygen, and as
1 volume of oxygen on combining with sulphur forms 1 vol-
ume of sulphur dioxide, which in turn requires ^ volume of
oxygen to form the trioxide, it is plain that 14 per cent, of
sulphur dioxide in the burner gas is the highest theoretical
percentage possible ; as each 14 volumes of sulphur dioxide
containing 14 volumes of oxygen requires 7 volumes of oxy-
gen to form sulphur trioxide, or 21 volumes of oxygen in
all, in which case the burner gas would contain the fol-
lowing:

Volumes of oxygen as sulphur dioxide 14

Volumes of oxygen to form sulphur trioxide. ... 7
Volumes of nitrogen 79

Total 100

In practice, however, even if pure sulphur is used to pro-
duce the burner gas, this percentage would not be practicable,
as no matter what process is used a certain excess of oxygen
is found necessary. This excess of oxygen is usually not
less than 5 per cent, and the proportions therefore are about
as follows:

14 SULPHURIC ACID § 27

Volumes of oxygen as sulphur dioxide 14.0

Volumes of oxygen to form sulphur trioxide. . 7.0

Volumes of oxygen excess 5.0

Volumes of nitrogen with the sulphur trioxide 79.0

Volumes of nitrogen with the excess of oxygen 18.8

Total 123.8

From which it is evident that even when burning brim-
stone or pure sulphur, the percentage of sulphur dioxide in
the burner gas should not exceed 11 per cent. Asa matter of
practice, 10 per cent, is rarely exceeded, as with less air sub-
limation of the sulphur is likely to take place unless great
care is used.

22. When the question is one of roasting the metallic
sulphides, it is evident that the matter is further compli-
cated, as oxygen (and with it nitrogen) must not only
be admitted to oxidize the sulphur to the trioxide and
to provide for the necessary excess, but also to oxidize
the metallic contents of the ore. The calculation will, of
course, be different for the various ores used, but it
may be stated in general terms that the burner gas pro-
duced when burning the metallic sulphides should range
from 5 to 8 per cent, of sulphur dioxide. A less percent-
age than 5 per cent, can only be used (on account of its
dilution with inert nitrogen) at the expense of a larger
and, therefore, more expensive plant; nor, with reasonably
well-constructed burners, need the percentage of sulphur
dioxide fall below 5 per cent, unless under very exceptional
circumstances.

23. Available Sulphur. — As all the raw material for
the production of burner gas contains varying quantities of
impurities, and as it is quite impossible, at the temperatures
existing in the various furnaces used in sulphuric-acid manu-
facture, to entirely desulphurize any of these raw mate-
rials— various percentages of sulphur remaining in the ash
or cinder — it is manifestly advisable to base figures relating

§27

SULPHURIC ACID

o 5©

to the process or yield upon the amount of sulphur actually
available or existing in the burner gas as oxides of sul-
phur SO^ or SOy The loss in the desulphurizing process is
estimated separately, and it is to this available sulphur that

all calculations will refer. Certain losses of
sulphur occur in the process of desulphurizing
by the escape of gas during charging and
discharging and the various manipulations
connected with the roasting. Losses also
occur by partially roasted ore passing through
the furnace; this is generally due to care-
lessness on the part of the burner men.
Other quite unavoidable losses are caused
by the temperature of the furnace being
insufficient to convert the sulphides of cer-
tain metals occurring with the iron pyrites
into oxides, they remaining in the cinder as
sulphates.

:5a

■

•^

tfi

•

"-N

"^^

•^

;«

•

.5t

•

u
3

O (0

(A VM

♦J ^
'u *^

>^ ?,

is <«

o s

«J a
c 3

o 5
a> d

24:. Sources of lioes of Sulphur In
Roasting* — As the metallic sulphides are
sold to sulphuric-acid manufacturers on the
basis of total sulphur contents, it is well, in
comparing the relative values of any ores, to
consider how much sulphur will be inevitably
bound in this way as sulphates in the cinder
and therefore, under no condition will be of
value to the manufacturer or available for
oxidation to the trioxide.

Table III is based on the assumption that
all these sulphides are converted to sulphates,
which is by no means the case.

25. The following illustrations show what
is meant by ** available" sulphur, the cause
of loss of sulphur in roastinjj^, and the relation

of this loss to the yield or output of acid and to the value

of any given ore to the manufacturer.

r \

16 SULPHURIC ACID § 27

Illustration 1. — An iron bisulphide of great purity, such as the
Aguas Tenidas in Spain, contains:

Iron 46.60^

Sulphur 63.15^

Silica m%

Arsenic \

^PP^'' >- Traces

Silver ( ^ ^^^^

Gold '

Such an ore with reasonable care can be roasted down so that the
cinders will not contain more than .5 per cent, of sulphur. As the cin-
ders will weigh only about 80 per cent, of the ore, the total loss of sul-
phur will be only .4 per cent., and this, so far as the metallic sulphides
are concerned, would seem in practice to be the irreducible minimum,
the manufacturer obtaining from the ore 52.75 per cent, of sulphur, or
99.24 per cent, of the sulphur for which he pays. In addition to this
loss, there will be more or less loss from the escape of gas from the
burners, varying with the excellence of construction of the burners
and the care exercised by the burner men. Other losses at the burners
will amount, say, in all to .6 per cent., making the total loss in ore
buying and roasting 1 per cent, of the sulphur contents, leaving 98.12
per cent, of the sulphur available for acid making, or as sulphur oxides
in the burner gas.

Illustration 2. — A Norwegian pyrites contains by analysis:

Sulphur 44.50 Lime 2.10

Iron 39.22 Magnesia 01

Copper 1.80 Oxygen, as /^«(?j . . .50

Zinc 1.18 Insoluble 10.70

It is plain that on a complex ore of this nature the loss in roasting

the bisulphide cannot be less than the loss in roasting the ore of the

previous example, or .5 per cent., in addition to which the ore contains

impurities that will hold sulphur in the cinders as sulphates, as stated

in the above table. The roasting losses will stand as follows:

Percentage
of Sulphur

Roasting loss on bisulphide of iron in cinders 40

Sulphate of zinc in cinders, 1.18 at .50 59

Sulphate of copper in cinders, 1.80 at .50 90

Sulphate of lime in cinders, 2.10 at .57 1.20

Sulphate of magnesia in cinders. .01 at .80

Sulphate of iron in cinders, .50 at .60 30

Total loss 3.89

In such an ore, therefore, the manufacturer will under no circum-
stances be able to obtain from the ore more than 41.11 pei cent, of the

§ 27 SULPHURIC ACID 17

sulphur, or 92.38 per cent, of the sulphur for which he pays. Adding
the further loss of .6 per cent, of sulphur in gas, etc. in the roasting
process, it brings his total loss up to 3.99 per cent, of sulphur, leaving
40.51 units of sulphur, or 91 per cent, of the sulphur, available for
acid making or as sulphur oxides in the burner gas.

In purchasing ore, it is further necessary to consider the
effect of low-sulphur contents on costs of freight, labor of
handling, and room taken up in the furnaces an(J storage
bins, etc. ; for instance, in purchasing brimstone, 1 per cent,
of these costs, at the outside, is on waste material, while
in dealing with an ore containing 50 per cent, of available
sulphur, 50 per cent, of these costs on the above accounts is
on waste material, and so on.

26. Yield and Method of Calculating Yield of Sul-
phuric Hydrate. — The possible theoretical yield obtainable
from 1 unit of actual sulphur, say 1 pound or 1 kilogram, is
2.5 pounds or kilograms of sulphur trioxide or 3.0025 pounds
or kilograms of sulphuric monohydrate H^SO^, which cor-
responds to 100 per cent, of either of the above products ; of
course, such yields are never realized in practice. A yield
of 98 per cent. (2.45 pounds of sulphur trioxide or 3.0013
of H^SO^ is probably the extreme average limit of even
the best-managed and best-constructed acid works, while
97 per cjent., or even 96 per cent., is considered extremely
good average work.

As these figures are based on actual chemically pure sul-
phur, the importance of the above remarks becomes evident.
At various factories, various and very loose data are used for
the calculation of yield. Some factories express their yield
in terms of sulphur shown by assay in the ore, without
reference 'to the loss shown by sulphur held as sulphates,
which can, under no circumstances, be recovered. Others
neglect the gas losses in the desulphurizing furnaces. The
safest way is to consider the available sulphur as that
actually contained in the burner gas as sulphur oxides; or
in practice to deduct from the assay value in sulphur of any
particular ore a suflScient percentage to allow for the inevi-
table loss in the cinder and gas at the furnaces.

SULPHURIC ACID

§27

^L'OQ= 'OS

1—1

<??

m'Z9^*os^//

•

<^^

'9fl s n oOfi

2ifi8-89 = 'OS

CO-

-<*

'7?

1—1

1^9 U = 'OS^'H

•

'^^

(T^

•99 S n o09

PPV "M

^11^99 - 'OS

»o

2>-

m'\% = 'OS*//

•

1—1

•

■

99 s n oZd

3
0

PPV J3AVOX-J3AO|f)

^88*9: = 'OS

'>*

^'26 = 'OS'^H

•

•99 S n .99

3
0

1— t

r-t

lou;iA JO no

• *

i!iHr6i= 'oy

CO
2>

1—1

^:6 = 'OS^H

»c

-Ti

•

1—1
CO

CO
CO

:^U86 = 'OS^'H

•

»c

p^iipsia

CI.

, :^9589"l« = 'OS

1— t

m^^

^)0l = V>.V7/

■

•

91BjpXqouoj^

^18 eft - 'oy

if}

»c

1—1

CO
CO

1 \^Z= 'OS )

•

1— t

•

X ;

"7^

im'm = 'OS

C
3

5^ 1

i ^9 = *0S*// I
'( ^Of = 'OS )

•

»c

'>?

•

• 1

'J*

'T^

;^i;9'o6= 'as'

S ^Of = 'OS""// )
1 ^9 = 'OS S

3
0

•

1— <
CO

;^oi = 'OS

T3
3

o
o

g 1

1 opupXijuv

•

■

»c

'ji

•

i- 1

ounqdins

CO 1

j [TMij^ijo^qx

»c i

JO ai<u;uaoj9(£

a-. ,

§ 27 SULPHURIC ACID 19

The same laxity is shown in estimating the consumption
of nitrate of soda used in the chamber process. At one
factory it will be expressed in terms of the actual available
sulphur; at another, on the sulphur assay of the ore used;
at another, even on the tonnage of pyrites burned. In the
case, therefore, of a factory using an ore containing 42 per
cent, of sulphur by assay, of which 38 per cent, only was
actually available, the percentage of sodium nitrate is actually
at different factories expressed as follows :

Sodium nitrate used per ton of ore 1.145^

Sodium nitrate used per ton of sulphur by assay 2.71^

Sodium nitrate used per ton of sulphur actually

available 3.00^

27. Table IV covers the yields, on actual sulphur, of the
principal solutions of sulphur trioxide in practical use.

The estimation of yield in a factory is not a very simple
matter. In fact, it is impossible to obtain the actual yield
except as the general average of a great number of observa-
tions and measurements extending over a considerable
period.

28, Sometimes every day, though usually once a week,
a record is taken of all the acid of various strengths in all
the apparatus and storage tanks of the system. The
dimensions of all this apparatus are usually tabulated, so
that every inch in depth corresponds to a certain cubic
capacity. The cubic contents of the acid of different
strengths being thus ascertained, all -these acids of
different ' strengths are reduced by Table I to one
strength — in a fertilizer factory, for instance, to terms
of 50° Baume; in other factories to terms of mono-
hydrate, or 66° Baum6.

In reducing these acids, careful note should be taken of
their temperature — although this is not usually done, the
error probably being considered as a constant one. In this
allowance for temperature, Table V is used by the Manufac-
turers* Association of the United States.

SULPHURIC ACID

§27

TABL.E V

Allowance for Temperature

At 10° Baum6
20° Baume
30° Baum6

40°
50°
60°
66°

Baiime
Baume
Baume
Baume

46. Fahrenheit
31.8 Fahrenheit
30.25 Fahrenheit
31.46 Fahrenheit
34.69 Fahrenheit
40.00 Fahrenheit
43.24 Fahrenheit

1° Baum6
1° Baume
1° Baum6
1° Baum6
1° Baum6
1° Baum6
1° Baume

From the total stock of acid obtained in this way is
deducted the amount on hand at the previous time of stock
taking; and the amount deducted from stock during the
period, either for use in other departments or sold, is added.
The result gives approximately the amount made during the
intervening period, and the yield is deduced either from
Table V or by calculation from the amount of sulphur used
during that period, a record of which has also been kept.
These records and measurements are usually taken by the
superintendent or acid maker and are checked by him at
intervals, together with some one of the proprietors or
general officers of the company. After a certain time, the
general average of the work done at any given plant can be
ascertained with fair accuracy so long as the same raw
material is used and the sulphur available has been deter-
mined with sufficient accuracy.

29. Sometimes the yield is roughly estimated, especially
in the contact process, by the difference in content of sul-
phur oxides contained in the burner and exit gases. The
formula given in Art. 53, for use in testing the burner gas,
enables this calculation to be made.

30. Another calculation that must often be made by an
acid maker is in connection with the mixing of acids of

§ 27 SULPHURIC ACID 21

various strengths in such a way as to produce an acid of any
desired strength. This is done by Gerster*s formula for
mixing a strong solution of sulphur trioxide with a wea,k
solution of sulphur trioxide to produce an intermediate
solution of sulphur trioxide of any desired strength. This
formula is as follows:

^=100^-^:1^, (1.)

a — c ^ '

when X = quantity of weak solution required to mix with

100 parts of the strong solution;

a = total sulphur trioxide in 100 parts of the solu-
tion desired;

b = total sulphur trioxide in 100 parts of the strong
solution ;

c = total sulphur trioxide in 100 parts of weak solu-
tion.

When the percentages of the solutions are given in terms
of monohydrate instead of sulphur trioxide, it is only neces-
sary to multiply the percentages of the monohydrate H^SO^
by .816326 to reduce them to terms of SO^ as mentioned in
Table I.

THE PRODUCTION OF SULPHUR
DIOXIDE OR BURNER GAS

31. General Remarks. — Commercial processes for the
manufacture of sulphuric acid are not intermittent, but con-
tinuous. It follows, therefore, that to secure regularity in
these processes all the separate factors must be as regular
and uniform as it is possible to make them. Furthermore,
the process consists of a series of chemical combinations,
which in any given plant are proportioned as to volume to
the size of that plant. If absolute regularity can be secured
in the various chemical combinations, then the maximum
work or output will be secured from such plant. Any
irregularities will result either in an incomplete series of

22 SULPHURIC ACID § 27

combinations and consequent waste of raw material, or a
reduction in output and consequent waste of capital outlay,
owing to incomplete utilization of the plant, or both.

The first requisite, therefore, in sulphuric-acid manufac-
ture is a uniform steady stream of gas of constant composi-
tion and volume. This gas should be produced at the least
cost for labor and repairs and with as complete an oxida-
tion of the sulphur in the furnace as possible. Unfortu-
nately, all furnaces, except some mechanical furnaces for
desulphurizing fines and one or two furnaces for burning
brimstone are intermittent in their action. That is, the
brimstone or ore is fed to them and the desulphurized cinder
discharged at intervals. It is only, therefore, by the most
skilful and careful work and attention to numerous details
that even an approximation can be had to the desirable con-
dition of the burner gas above referred to.

32. To counteract the intermittent character of the
individual furnace, the following points must be observed:

1. A considerable number of furnaces of small capacity
are used, and are charged and discharged in series. For
example, a sulphuric-acid works is designed to oxidize in
24 hours 14,000 pounds of actual sulphur to the trioxide. The
ore available is iron pyrite in lump form, containing 50 per
cent, of sulphur, about 1.5 per cent, of which will be lost
either by being retained in the cinders or on other accounts,
making the ore contain 48.5 per cent, of available sulphur;
28,800 pounds of ore will be required daily. To roast this
ore, twenty-four burners will be used, each having a capacity
for roasting 1,200 pounds in 24 hours. This ore is charged
to each furnace in two charges of 600 pounds each, one
every 12 hours, or the whole charge is divided up into forty-
eight charges of GOO pounds each, so that one furnace of
the twenty-four will be charged with 600 pounds of ore
every half hour during the 24 hours. Furnaces are selected
to be charged in rotation in such a way as to preserve as
nearly as possible even conditions in every part of the
bench of burners.

§ 27 SULPHURIC ACID 23

2. The furnaces are so constructed that the amount of
air admitted to each furnace may be under as complete con-
trol as possible. It is evident that when any furnace has
received a new charge of ore containing 50 per cent, of sul-
phur, it will require more air than it will (5 hours later, when
much of the sulphur is burned off, and still more than it will
when the sulphur is almost entirely burned off. In fact, the
admission of much air when the ore only contains a small
percentage of sulphur merely tends to cool the furnace and
so prevent the thorough roasting of the ore. By judiciously
regulating the admission of air to the individual burners,
the general average of the gas in the flue common to all the
burners is kept reasonably strong: and by the subdivision
of the whole charge as above, a general average of gas is
maintained that is as near an approach to continuous, uni-
form work as is possible under the circumstances.

FURNACES^ AND BTTRKKBS FOR THE PRODUCTION

OF BITRXER GAS

33. General Remarks. — In the production of the burner
gas and the efficients desulphurization of the various raw
materials by the different furnaces now to be described —
while various points peculiar to the management of each
furnace will be pointed out — nothing but actual experience
will secure satisfactory results. The minutest details tend-
ing to secure regularity must be insisted on in the manage-
ment of the furnaces. Each ore or material has its own
peculiar behavior in the furnaces, which when understood
must be attended to.

When natural draft is used, meteorological conditions
must be constantly considered and the drafting of the fur-
naces modified accordingly. Much trouble and anxiety is
saved in this respect by the use of fans, or other apparatus
devised to make the draft positive and controllable. Above
all, it must be kept constantly in mind that the desired

SULPHURIC ACID

§37

§ 27 SULPHURIC ACID 25

object is a uniform stream of burner gas of constant com-
position and volume, with as complete desulphurization of
the ore as may be possible.

If the further oxidation of the sulphur dioxide of the
burner gas to the trioxide is to be carried out by means of
the chamber process, it is necessary to mix the burner gas
at this point with nitric-acid fumes. The nitration of the
gas will only be mentioned here in so far as it forms an
adjunct to the desulphurizing furnaces; in other words,
when the nitric acid is supplied by the decomposition 'of
sodium nitrate and sulphuric acid by the heat of the burner
gas. This method of adding the nitric acid is called /^///V/^,
as it is done in large cast-iron pots, placed in a chamber or
enlargement of the main gas flue of the furnaces, which is
called the niter oven.

BRIMSTONE BURNERS

34, One of the simplest forms of brimstone burners is

shown in Fig. 1 {a) and (b)^ (a) being the plan showing sev-
eral burners connected with the common flue /r, and {b) a
vertical section on the line x y. The sulphur is charged
through the door ;;/ upon the cast-iron pan a^ where it is
burned. The supply of air to the burning sulphur is care-
fully regulated. The gases containing the sulphur dioxide
collect in the chamber b and pass through the flue d into
the common flue h. To prevent overheating pan a^ air is
admitted imder it through ;/, passing through r, e^f, and ^,
where it finally mixes with the gases and sublimed sulphur
coming from //, as shown in {a). These mixed gases now
pass into the combustion chamber /, where the combustion
is completed.

If the burner gas is to be used for making sulphuric acid
by the chamber process, it is mixed with fumes of nitric
acid evolved in the pot/ by the action of sulphuric acid on
niter. The gas now passes through the flues k and / to
the Glover tower. If it is to be used in. the contact proc-
ess, the nitrating is omitted.

U SULPHURIC ACID § 37

35. Ilftri'lson-nialr Itrlmstoiie Itiimer. — This burner,

which is of the continuous-feed and intermittent-discharge
type, is shown in plan and longitudinal section in Fig. 3 («)
and (d). The brimstone ,is fed into the burning pan a
through the funnel if, which is kept full. The brimstone
settles as fast as that on the pan melts. Air for the com-
bustion is supplied through the door c. The sulphurous
gases pass from the chamber d through the flues e and /"

into the flue or chamber i; additional air for completing the
combustion is supplied to / through the openings/'. In
passing from / to t, the gases are led through a series of
baffle walls lined with pigeonholed brick, shown at ^, and
over the niter pots /i. From the flue t the nitrated gases
pass through the flues / and ^ to the Glover tower.
About once in 24 hours the ashes are removed through
the door c.

§ 27 SULPHURIC ACID 27

PTRITES BURNEU8

36. Furnaces, or burners, for roasting pyrites for the
recovery of the sulphur, as 5C^„ are of many styles, depend-
ing both on the nature of the pyrites used and on the man-
ner of their operation. The following descriptions will
serve to illustrate the most important forms.

37« Falding Xiump Burner. — This furnace is used for
self-roasting ores. It has an intermittent feed and is oper-
ated by hand. A bench of six furnaces is showji in detail
in Fig. 3 (a), (^), (r), {d), and (r). {a) is a side elevation,
also showing vertical sections through several parts; (6) is a
plan showing horizontal sections through several different
parts; {c) shows a vertical section from front to back through
the center of an individual furnace; {(/) and {e) will make
themselves clear in the following description:

38. At a is the grate upon which the ore is burned. The
thickness of the bed of ore carried on the grates will be from
2 to 2^ feet, as shown in (^), but will vary somewhat accord-
ing to the sizing and character of the ore. It must in any
case permit a passage of the air uniformly through its mass
and not in spots or against the furnace walls. The ore is
shoveled into the furnace through the charging door ^, and
must be spread as evenly over the surface of the bed as
possible, being slightly deeper against the walls of the fur-
nace, as shown in (c). The grate bars a are usually bars of
square wrought iron from IJ^ to 2 inches square, slightly
rounded where they pass through the supporting cast-iron
bearers ei^^ a^, a^. These bars must be spaced so as to be
best adapted for the size of the ore to be burned. When
the ore is sized by screens, it is a good plan to have a certain
number of furnaces spaced to accommodate each size of ore.
If the grates are spaced too closely, the larger lumps will
not pass through and the draft will soon be seriously inter-
fered with; if they are spaced too far apart, the bed will
drop through too rapidly and be difficult to control. At the
front of the furnace the bars pass through a wrought-iron
plate c and c\ that can be removed in two sections; as the

28 SULPHURIC ACID § 27

bars are turned down to a circular section where they pass
through the plates, the plates fit closely and prevent the
entrance of ** false'* air into the furnace; the plates also
tend to steady the grate bars. In order to drop the roasted
ore through the grate, use is made of a large wrench, or key,
fitting on to the square end of each grate bar. Each grate
bar is by this means twisted backwards and forwards a
few times, until an amount of roasted ore has been dropped
through the grate into the cinder pit i equivalent to the
amount of ore about to be charged through the charging
door b. The roasted cinder is removed by means of the
door d'. The cinder-pit door d' is provided with a slide or
gate valve d for regulating the admission of air for com-
bustion.

39. The furnace illustrated has hollow front walls ^,
which serve to prevent radiation of heat from the furnace,
permitting the burner gas to be passed to the Glover tower
at a high temperature; or, if desired, permitting the air
supply for combustion of the ore to be preheated. The air
passes through the regulator h into the air duct/, hollow
tiles e, and side channels g^ beneath the furnace grate.
The door y is used for inserting a bar in case of bad clinker-
ing low down in the bed. The bricked-up opening k enables
the flue / to be cleaned when obstructed with flue dust.
Each furnace has a roof ///, in which is an opening n, through
which the burner gas finds its way into the main flue /
formed by the longitudinal arch ^, and thence into the flue p
common to both sides of the furnace bench, whence it is
carried by the cast-iron pipe q to the Glover tower. In case
nitration by means of potting is to be used, the niter oven
will be placed in /, which also serves as a dust collector for
coarse flue dust that may be carried over.

40, To start this furnace, about 2 feet of roasted ore is
put upon the grates. (Incompletely roasted ore and brick-
bats broken so that they can be passed through the grates
may be used.) A light wood or coke fire is then lighted on

30 SULPHURIC ACID § 27

the bed of each furnace. When the furnace is heated to a
dull-red heat, the coke or wood ash may be removed and ore
charged and the gas turned into the acid plant.

41. Maletra-Faldlng: Ftimace. — This type of furnace
is adapted to the roasting of fines. It is worked by hand
and has an intermittent feed. Fig. 4 (a) and (/;) shows two
sectional views of it, corresponding parts being lettered alike
in both views.

Ore is introduced by means of the hopper and bell a on to
the back end of the upper shelf d. These shelves can be
constructed either with fireclay slabs r, r, r, or brick arches,
as shown at </, ^/, d. After the furnace has been brought to
a red heat ore is spread over all the shelves. After the
expiration of a certain time the ore is raked off the lower
shelf r through the door/. The ore from shelf g- is then
pushed through the opening // on to shelf i\ through the
door /, and spread over shelf r. The ore from shelf j is
raked forwards on to shelf g^ over which it is spread. The
ore from shelf k is pushed through the opening / and spread
on shelf y, and so on until the ore from shelf d is finally
raked through the opening ;;/ by means of door*;/ and spread
on shelf o. Every shelf in the furnace is now covered with
ore except the upper shelf /?. A charge of ore is now intro-
duced through the hopper and bell a on to the upper shelf i
and spread over it. The furnace is now left for from
G to 12 hours, when the lower shelf is discharged and the
whole operation repeated.

These furnaces are generally constructed in groups of
from four to sixteen, each of which has a capacity of from
1,000 to 2,000 pounds of pyrites in 24 hours.

42. Herreshoff Furnace of the MaeDoiij?all Type.

This furnace is provided with mechanical rotary stirrers
and has a continuous feed. It is designed for burning fines,
and is illustrated in Fig. 5 (a), (^), and (r).

This furnace has five shelves, or hearths, a, a^ a, by b.
The rotating, central, hollow, cast-iron column c carries a

□P

•A

1^4

§ 27 SULPHURIC ACID 31

pair of arms on each shelf. The column and arms adjacent
to the column are kept safely cool by a current of cold air
drawn through the holes d by natural draft created by
the 30-foot stack e. The teeth, or stirrers, operating on
hearths a, a, a are set at such an angle as will gradually
work the ore from the center to the periphery, where it
falls through ports / on to the hearths b, b and finally
through discharge port/', which is closed by a balance valve
suitably weighted to control the discharge. The teeth on
the stirrers, operating on hearths by b, are set at such an
angle as to gradually work the ore from the periphery to the
center, where it falls through the annular ports g. The ore
is fed to the furnace by means of the hopper and plunger
feed A, h\ which is operated by the central revolving
column c in such a way as to supply a desired quantity
of ore at each revolution of the stirrer arms. These
usually make a complete revolution once in 2 minutes
and can be taken out and replaced by means of the
doors f, i„ /„ /„ i^,

43. The air for supplying the necessary oxygen for the
combustion of the ore and the production of a suitable
burner gas is admitted on the lower shelf through gate
valves ky ky etc., and passing through the furnace over the
roasting ore, finally leaves the furnace as burner gas through
the cast-iron pipe /, which connects with the common, or
main, cast-iron burner-gas flue ;//.

If the mechanical construction of this furnace has been
properly attended to, it can be readily started by first
removing the arms, covering the hearths with a bed of
roasted ore or cinder, then heating it to a dull-red heat by
means of light wood fires on each shelf, then replacing the
arms and feeding the ore in the usual way.

44« Spence Reciprocating Typo of Furnace. — This
style of furnace is designed for self-roasting fines. It has a
continuous feed, the ore being carried from one shelf to
another by means of a reciprocating device. Its mechanism is

SULPHURIC ACID § a?

shown in Fig. 6. The
exit flue a is for
the burner gases.
The hopper d is kept
full of ore, which is
fed into the furnace in
front of the rake i up-
on the hearth c by the
continuous feed d'.-
As the rake i is drawn
over the hearth c
by tlie reciprocating
mechanism shown at
the right of the fig-
ure, it draws and
spreads the accumu-
lated ore over the
„ hearth and causes
g part of it to fall
through the open-
ing c' upon the next
lower hearth d. On
its return stroke the
rake on hearth </ car-
ries this ore over the
hearth in the oppo-
site direction, it then
falling through the
opening d'. From
the lower hearth / it
falls through/' into
the pit ff. A similar
operation takes place
on each hearth.

45. The ore rakes

have triangular cast-
iron teeth and extend

§ 27 SULPHURIC ACID

34 SULPHURIC ACID § 27

across the furnace ; they are attached to the skids / '. The
reciprocating mechanism consists of a hydraulic cylinder ;/
with a piston and piston rod m. The motion of the piston
is transmitted to the rakes by means of the rods / and y and
the frame k. Water is furnished to either end of the cylin-
der through the pipes / and q by the pump O by means of
automatic valves.

When the furnace is to be started up, a fire is built in the
fireplace A, but after the furnace is well under way it is
bricked up, the sulphur of the ore being the only fuel
necessary for the continuation of the operation.

46. Rlienania Muffled Type of Furnace. — This fur-
nace is for roasting refractory ores ; the feed is inter-
mittent. It is illustrated in Fig. 7 (rt), (^), and (r). ^, ^, and c
are the hearths, or shelves, upon which the ore is burned.
Ore is fed upon the hearth a from time to time through
the openings d. The ore is worked from shelf to shelf by
means of slice bars introduced through the doors ^, and
finally passes through the ports / to the cinder pits g^ from
which it is periodically removed.

Heat for combustion is supplied by the fireboxes //; this
heat passes through the whole length of the double furnaces
and returns by means of the flues /, passing into the stack
at j\ The products of combustion of the ore pass over
the ore and in the opposite direction into a flue ^, on the
top of the furnace, and thence by dampers / into the
stack m connecting with the Glover tower.

These furnaces are built in blocks of four, fired by means
of two fireboxes with two doors each, at one end of the block
of four; the fire flues / pass from end to end of all four fur-
naces and back again

47. MacDougall Type of Muffled Furnace. — This
furnace closely resembles in structure and operation the
Herreshoff furnace already described. When these fur-
naces are muffled, that is, supplied with separate and distinct
combustion chambers and flues for the purpose of intro-
ducing heat into the furnace other than that produced by

SULPHURIC ACID

n n D

36 SULPHURIC ACID § 27

the combustion of the ore itself, .and the products of com-
bustion are kept separate and distinct from the burner gas,
they are constructed in groups of four. The feed is con-
tinuous. The fuel used may be oil, natural, or producer gas.
Fig. 8 (a) and (d) shows this furnace in plan and vertical
section. The revolving arms j are operated in the same
manner as in the Herreshoff furnace. Ore is fed in through
the hopper // to the upper hearth and gradually worked out-
wards by the revolving arms and down to the next lower
hearth, then towards the center and through another opening
to the third hearth, etc. The fuel is supplied by means of
the pipe/ and branches ^ to the combustion chamber r, and
thence to the flues s around the muffles. The products of
combustion pass off through the stacks / and the burner gas
through the pipes k.

TESTING THE BURNER GAS

48. Collecting the Sample. — Gas is aspirated from the
flue common to all the furnaces at a point chosen so as to
secure a reliable average of the gas. If there is any doubt
as to the gas being an average, it should be aspirated at
several points until a point yielding a satisfactory average
is obtained.

49. Reich's Test for Sulphur Dioxide. — This test,
which is described fully in Quantitative Analysis^ is gen-
erally used, but is modified for sulphuric-acid works as
follows : The deci-normal solution of iodine, containing
12.r.5 grams of iodine per liter, and the starch solution are
pre[)ared in accordance with the instructions given in
Quantitative Analysis. The solutions must be kept in a
dark, cool place. It is a good precaution to use small bot-
tles, holding just sufficient for the day's tests, for use in
the works, leaving the stock in the laboratory.

As in practice it is often necessary to make tests in
several different parts of the works, it becomes necessary,
for making the test, to fit up a simple cheap apparatus that

§ 37 SULPHURIC ACID 37

can be left at each place. The apparatus shown in Fig. 9
can be r^dily put together and arranged on a rough shelf
or shelves, in almost any place, so as to be ready for instant
use. A pipette being left on the shelf, it is only necessary
to carry around the two small bottles of solutions. A spare
3-Iiter jar being kept on the shelf, the aspirating water
can be saved, or in case many tests must be made and

water-supply pipes are available, a water aspirator can be
used; failing this, a steam aspirator. In some works the
pipes through which the gas is aspirattd are all led to the
laboratory, and being supplied with valves and a steam
aspirator, tests can be made in the laboratory without going
to the different parts of the works. This is a very con-
venient though somewhat costly plan.

SO. This apparatus consists of a 1-inch iron pipe^ pene-
trating the flue A, the gas in which is to be tested. This pipe
is connected by means of the rubber tubey"and pinch cock e
with the absorption bottle d. The glass tube / penetrates

38 SULPHURIC ACID § 27

the rubber stopper and extends nearly to the bottom of
the bottle. The exit tube / extends just through the stop-
per. By this arrangement, the gas drawn through / bubbles
through the reagent in d. The rubber tube k connects d
with the 4-quart aspirating bottle a. In construction, this
aspirator is similar to the absorption bottle, the positions
of the tubes being reversed. When filled with water, th •
latter is siphoned off through / and ;//, creating a partial
vacuum in the upper part of a. During operation, the
volume of water (which is equal to the volume of gas used)
drawn from a is measured in the graduated cylinder n,

51. To make the test, fill the large bottle a with water;
see that the stopper is perfectly tight ; start the siphon by
slight suction through the nozzle b and close the pinch cock^;
fill the 8-ounce bottle d about one-quarter full of clean water
(slightly warmed in winter); and pour into this about a
teaspoonful of the starch solution. Then, by means of the
pipette take 10 cubic centimeters of the deci-normal iodine
solution and add to the water and starch solution in the
8-ounce bottle ; replace the rubber stopper tightly and close
the pinch cock e between the flue and the small bottle. Then
open the pinch cock c at the nozzle and allow the water to
waste; when the water ceases to run, proving the tightness
of corks and connections throughout the apparatus, open
the pinch cock e between the flue and the small bottle d.
Take the small bottle in the left hand, keeping the right
hand on the pinch cock c at the nozzle, and shake the bottle
not too violently, holding it to the light in such a way
that any change in color can be readily noted ; when a con-
siderable change occurs in the color, stop the flow of water
with the right hand. If the color does not entirely disap-
pear, aspirate a little gas carefully until it does ; then close
the pinch cock e between the flue and the small bottle. The
tube / between the flue and this pinch cock is now filled
with the gas to be tested. Remove the stopper from the
small bottle and add 10 cubic centimeters of the iodine
solution with the pipette; replace the cork tightly; open the

§ 27 SULPHURIC ACID 39

pinch cock nearest the flue; then, with the pinch cock c in
the right hand carefully waste water until the liquid in the
glass tube, terminating the tube from the flue, is depressed
to the bottom ; or, in other words, until the tube is filled with
the gas to its extreme end in the small bottle. Just before
the first bubble of gas would escape and pass through the
solution, allow the water to commence running into the
graduated measuring jar ; shake the small bottle as before and
stop the water running the instant the color is discharged.
The number of cubic centimeters of water that the jar holds
at the point of discharge of color from the solution repre-
sents the volume of gas required to decolorize the solution,
from which the percentage of sulphur dioxide in the burner
gas is calculated.

, 62. The reaction talcing place is as follows:

2/+ 5(9, + 2//,(9 = %HI-\- H^SO,

Omitting any correction for temperature and pressure,
the percentage of sulphur dioxide is calculated by means of
the following formula:

^"~w + l.lUx;/* ^ ^^

/ = percentage of SO^,
n = the number of cubic centimeters of ^^ normal iodine

solution used;
tn = the number of cubic centimeters of water run into

the measuring jar.

If the percentage of sulphur dioxide in the gas is very
small, and, thus, ;// is very large in proportion to //, the
formula may be simplified into

111.4 X« .ov

In testing exit gas, using 10 cubic centimeters of a jj^
normal or centi-normal solution of iodine, formula 2 becomes

^ = ;;/+!. 114' ^^'^

40 SULPHURIC ACID § 27

As in the case of formula 3, when the percentage of SO^ is
very small and ;;/ is very large in proportion to n (as it
usually is), the formula becomes

^ in ^ '

By means of the following tables, the percentage of sul-
phur dioxide in burner and exit gases can be read directly
from the volume of water in the measuring jar.

53. To calculate the yield of sulphur dioxide from the
difference in content of SO^ in the entering and exit gases,
the following formula is used :

Yield = f "" ^^ \, (6.)

where a = the percentage of SO^ in the entering gas;
b = the percentage of SO^ in the exit gas.

CALCULATTOX OF VOI^I^TVIE OF BUUXER GAS

54. The general principles contained in Theoretical
Chemistry regarding corrections for temperature and pres-
sure and the corrections of gaseous volumes treated in
Physics, Theoretical Chemistry, and Quantitative Analysis
apply to burner gas. A rough approximation of the volume
of burner gas at 0° C. and 960 millimeters barometric pres-
sure, or at 32° F. and 29.92 inches barometric pressure may
be made as follows:

55. One liter of sulphur dioxide weighs 2.8()330 grams, "or
1 cubic foot weighs .1787 pound. Therefore, 1 pound of
sulphur burned in 24 hours produces 11.191908 cubic feet
of sulphur dioxide, or .0077722 cubic foot per minute; there-
fore, neglecting the sulphur trioxide formed,

(.77722 X actual available sulphur burned in 24 hours) _
average per cent, of SO^ in burner gas produced ~~

the cubic feet of burner gas per minute;

§27

SULPHURIC ACID

TABLK VI

TABLE FOR FINDING TIIK PKIUENTAGE OF SO, FN BURNER

GAS WHEN USING lO CUBIC C'ENTIMETEItS OF DECI-

NORMAL IODINE SOLUTION (CALCULATED

BY FORMULA «)

Cubic

Per

Centi-

Per

Centi-

Per

Centi-

Per

Centi-

Cent.

meters

Cent.

meters

Cent.

meters

1 Cent.

meters

of SO,

of .sy;, '

of .v^;-,

, of SOi

Water

Water
337 . 0

•
1

5.5

Water

t 7.8

Water

« • • •

3.2

191.0

132.0

1.0

1,103.0

3 . 3

327.0

5.6

188.0

7.9

130.0

1.1

1,002.0

3.4

317.0

5.7

184.0

8.0

128.0

1.2

917.0

3.5

307 . 4

5.8

181.0

8.1

126.0

1.3

846 . 0

3.6

298.0

5.9

178.0

' 8.2

125.0

1.4

785.0

3.7

290.0

6.0

175.0

8.3

123.0

1.5

732 . 0

3.8

282 . 0

6.1

172.0

8.4

122.0

1.6

685 . 0

3.9

275 . 0

6.2

169.0

' 8.5

120.0

1.7

644 . 0

4.0

267.0

6.3

166.0

8.6

118.0

1.8

608.0

4.1

261.0

6.4

163.0

8.7

117.0

1.9

575.0

4.2

254 . 0

6.5

160.0

8.8

116.0

2.0

546 0

4.3

248 . 0

6.6

158.0

8.9

114.0

2.1

519.0

4.4

242 . 0

6.7

155.0

9.0

113.0

2.2

495.0

4.5

236 . 0

6.8

153.0

9.1

111.0

2.3

473 . 0

4.6

231.0

6.9

150.0

9.5

106.0

2.4

453.0

4.7

226.0

7.0

148.0

10.0

100.0

2.5

435 . 0

4.8

221.0 1

7.1

146.0

10.5

95.0

2.0

417.0

4.9

216.0

7.2

144.0

11.0

90.0

2.7

402 . 0

5.0

212.0 i

7.3

142.0

11.5

86.0

2.8

387 . 0

5.1

207.0

7.4

139.0

12.0

82 . 0

2.9

373.0

5.2

203.0

7.5

137.0

3.0

360.0

5.3

199.0

7.6

135.0

3.1

348.0

5.4

195.0

7.7

1 34 . 0

SULPHURIC ACID

§27

or, letting x
a

b =

then,

cuj)ic feet of burner gas per minute at 0° C.*;
available sulphur in pounds burned in
24 hours ;

average percentage of sulphur dioxide in the
burner gas produced ;

.77722 X a
b ' '

TABI.B VIT

;r =

(7.)

TABL.B FOR FINDING THK PERCENTAGE OF SO, IN BXIT GAS

WHEN USING lO CUBIC C ENTIMETERS OF CENTI-

NORMAL IODINE SOLUTION (CALCULATED

IIY FORMUI4A 4)

Cubic

Per

Centi-

Per

Centi-

Per

Centi-

Cent.

meters

Cent.

meters

Cent.

meters

of 50,

Water

•

.05

2,226 . 9

.55

201.5

1.05

105.0

.10

1,112.9

.60

184.6

1.10

100.1

.15

741.6

.65

170.3

1.15

95.8

.20

555.9

.70

158.0

1.20

91.7

.25

444.5

, .75

147.4

1.25

88.0

.30

370 . 2

.80

138.2

1.30

84.6

.35

317.2

.85

130.0

1.35

81.4

.40

277.4

.90

122 . 6

1.40

78.5

.45

245 . 5

.95

116.2

1.45

75.0

.50

221 . 7

1.00

110.3

1.50

73.2

Cubic

Per

Centi-

Cent.

meters

of 50,

Water

1.55

70.8

1.60

68.5

1.65

66.4

1.70

64.4

1.75

62.6

1.80

60.8

1.85

59.1

1.90

57.5

1.95

56.0

2.00

54.6

For example, 10,000 pounds of available sulphur is burned
in 24 hours with the production of a burner gas containing
7.5 per cent, of sulphur dioxide, substituting in formula 7,

.'77722 X 10 7,772.2

X =

7.5 7.5

= 1.036.3 cubic feet of burner gas per minute, containing

7.5 per cent, of sulphur dioxide.

§ 27 SULPHURIC ACID 43

This quantity of gas at normal pressure is corrected for a
temperature of, say, 00° F. or 15.6° C. by the following
formula:

y. = K + y^. (8.)

F, = volume at the given temperature;
F« = volume at 0° C. ;

/ i= temperature in degrees C. at which volume is to be
calculated.

Substituting, V^ = 1,036.3 cubic feet, / = 15.6° C,

V, = 1,036.3 + •^^!; = 1.095.5 cubic feet per

mmute.

The approximate temperature at which the burner gas
leaves the common flue of a bench of burners is about
1,000° F., or, say, 538° C, and at this temperature the vol-
ume of burner gas produced as above would be 3,078.6 cubic
feet per minute. With gas containing only 5 per cent, of
sulphur dioxide, the quantities at 0° C. and 538° C. would be,
respectively, 1,554.4 and 4,617.6 cubic feet per minute.

56. It is evident, therefore, that great economy in the
size of the apparatus is effected by keeping the gas as
strong as possible; and also that in order to prevent
unnecessary obstruction in the flues and apparatus, atten-
tion must be given to designing them of suitable capacity,
to handle the volumes of gas in accordance with the approxi-
mate temperature of the gas at the different stages of the
process.

THE CATALYTIC, Oil CONTACT,

PROCESS

57. Preliminary llemarks ConeerniiiK Contiu't Phe-
nomena.— In considering the so-called contact phenomena
in chemistry, it must not be forgotten that contact is a neces-
sary condition for every chemical reaction. Other conditions
remaining constant, the rate of progress of a chemical

44 SULPHURIC ACID § 27

reaction is accelerated by increasing the number of points of
contact. To insure complete reaction between solids, it is
necessary to reduce them to very fine powder and to mix
them as thoroughly as possible. These considerations may
throw some light on the large class of contact reactions;
that is, such as appear to proceed from the mere presence of
certain special substances. Porous or powdery substances
are very prone to act in this way, especially spongy or very,
finely divided platinum and charcoal. A number of other
substances, such as finely divided silica, act in a similar way.
Another consideration is the action, by contact, that two
substances rich in oxygen have upon each other, in that so
long as they are separate they retain their oxygen ; but upon
contact oxygen is liberated from both of them. As, for
example, a solution of bleaching powder, which does not
evolve oxygen when heated by itself, but upon the addition
of a small quantity of certain oxides, for instance, cobalt
oxide, first oxidizes the cobalt oxide to a higher oxide, which
in contact with the bleaching powder decomposes into oxygen
and the lower oxide. This resulting lower oxide, on contact
with the bleaching powder, again results in the higher oxide,
which again gives up its oxygen and produces the lower
oxide, and so on

68. The action of nitrogen oxides in the chamber process
is noteworthy as showing that intermediate forms of reaction
may be found in the contact, or catalytic, phenomena. In
this case a small quantity of nitrous oxide induces a definite
chemical reaction between large masses of sulphur dioxide,
oxygen, and water, forming sulphuric acid, the N^O^ being
finally again liberated, as will be seen when considering the
chamber process.

In the case of the combination of sulphur dioxide and
oxygen by contact action, it is possible that either on account
of an electrical action induced by the contact, or for some
other obscure cause, a polarization or increased activity of
the oxygen in the air is procured, enabling it to combine
with the sulphur dioxide.

§ 27 . SULPHURIC ACID 45

59, Richter suggests that as all bodies having a high
heat of formation, and also those being decomposed at a high
heat, must have their heat of formation removed or con-
ducted away in order that their production may be at all
possible ; the catalytic action of many metals, for example,
platinum, in this reaction, may be due to their conducting
off the heat; or else that the bodies in question forming a
galvanic chain, the chemical energy is removed as electricity,
just as in the union of hydrogen and oxygen at ordinary
temperatures due to the formation of a polarization current.

60. Contact Mtiss or Material l"so<l In the Manu-
facture of Sulphuric Acid by the Contact Pi-ocess.

Broadly speaking, there are four contact masses in com-
mercial use for the manufacture of sulphuric acid, viz. :
(1) Asbestos, clay, pumice, or other porous material impreg-
nated or coated with platinum. (2) Porous or fibrous
material as above impregnated with cupric sulphate (blue
vitriol). (3) Mass composed of crusts formed of an earthy
or alkaline water-soluble salt impregnated or coated with
platinum. (4) Ferric oxide (roasted pyrites).

For the first class of contact masses, where the platinum
is combined with a fibrous or porous material insoluble in
water, there are two principal methods of preparation, the
first being to add finely divided platinum (platinum black),
previously prepared, to the fibrous or porous material ; and
the second, to add either a dry or liquid salt of platinum to
the inert material and then subject the mixture to a process
that will reduce the platinum.

61« The usual methods for preparing the first class of
contact masses are as follows:

1. The powdered fibrous or porous material is mixed with
platinum black, a combustible material required to secure
porosity (flour, bran, sawdust, cork dust, etc.), and an
agglutinative substance (gelatine, gum, etc.).

2. The fibrous or porous material is mixed with an oxide
or dry salt of platinum, a combustible material, and an
agglutinative. It is then dried and reduced by calcination.

46 vSULPHURIC ACID § 27

3. The fibrous or porous material is soaked in a platinum-
salt solution, reduced by one of the methods described in the
paragraphs immediately following, and after the addition of
the combustible organic matter and agglutinative, is molded,
dried, and calcined.

4. The fibrous or porous material is first impregnated
with a platinic chloride and then reduced by one of the
following-methods: (a) By plunging the saturated material
into a solution of ammonium chloride, ammonium-platinic
chloride {NH^J^Cl^ is formed. The whole is then dried
and calcined, {b) By plunging the material into a bath con-
sisting of an alkaline solution of sodU and of platinum
chloride containing sufficient sodium formate to reduce the
platinum, evaporating, washing, and drying, (c) The mate-
rial saturated with platinum salts can be dried and submitted
to the action of hydrogen or of gas rich in hydrogen, such
as ordinary illuminating gas or even of hydrocarbon com-
pounds, (d) The following methods for the preparation of
platinum black may also be used.

63. Platlnuin black or finely divided platinum can be
made as follows; (a) Platinic chloride /^/^/^ is treated in a
concentrated potash lye with alcphol. The resulting powder
is washed successively with alcohol, hydrochloric acid,
potash, and water, {b) Platinum sulphate can be reduced
by alcohol, (r) By the calcination of a platinic chloride, as
calcium-platinic chloride CaPtCl^^ or ammonium-platinic
chloride {NH^^PtCl^. (d) By precipitating platinic chlo-
ride with zinc, (r) By heating an ammoniacal salt of
platinum, mixed with shreds of cork, in an open crucible.
(/) ^y t^^^ reduction of platinic chloride with admixture of
sodium carbonate, sugar, etc. {^) If 50 grams of platinic
chloride be dissolved in GO cubic centimeters of water and
70 cubic centimeters of a 40-per-cent. solution of formalde-
hyde be added, the mixture cooled, and then a solution of
50 grams of sodium hydrate in 50 grams of water added, the
platinum is precipitated. After washing with water, the
precipitate passes into solution and forms a black liquid

§ 27 SULPHURIC ACID 47

containing soluble colloidal platinum. If the precipitated
platinum be allowed to absorb oxygen on the filter, the tem-
perature rises 40° C. and a very porous platinum black is
obtained that vigorously facilitates oxidation.

Instead of the second class of contact material, some
manufacturers use cupric sulphate at a red heat as contact
mass. The salt is mixed into a paste with finely ground
clay, molded into the desired shape, and dried.

63. In the third class of contact masses (under the
Schroeder-Grillo patents), instead of the solid or integral
insoluble bases above referred to, use is made of the soluble
salts of the alkalies and of the alkaline earths, and of the
heavy metals, which salts, for the production of the contact
mass, are dissolved in water and then mixed with a solution
of the finely divided platinum salt, especially platinic chlo-
ride. It can be used in a solution so diluted that in
100 parts of the salt, serving as base or vehicle, less than
1 part of platinic chloride is sufficient. Even contact bodies
of .1 per cent., and less, of platinum contents are very
efficacious. This mixture of solutions is then evaporated
and the resulting salt crusts dried and broken up to about a
uniform granular size. The powder that is formed in this
reducing, or breaking-up, operation is dissolved afresh in
water and treated as before until all the material has been
converted into uniform granular size. The reduction of the
metallic platinum in the finest subdivision between the
molecules of the salts serving as vehicles for the platinum
takes place automatically upon heating. In practice, the
salts are always sulphates.

The technical advantage of this contact mass lies partly
in the simplicity of its preparation; in its activity, on
account of the extremely fine division of its platinum ; and
on the relatively small quantity of platinum required, both
because of its fine division and because the base used also
possesses catalytic activity. It is also regenerated readily
and the platinum can be easily and completely recovered,
on account of the solubility of its base, or vehicle, in water.

SULPHURIC ACID

§ 27 SULPHURIC ACID 49

When ferric oxide, the contact mass of the fourth class, is
used, it is in the form of pyrites cinders (desulphurized iron
pyrites), and these cinders must be porous and fresh. One
advantage claimed for this mass is the removal of the arsenic
from the burner gas in its passage through the cinders. It
is also necessary to dry the air supplied to the roasting fur-
naces and to dilute the gas with further admissions of dry
air after combustion and before it passes through the contact
mass of cinders.

64. Frasch Converter. — A further elaboration of this
process is the Frasch converter, which serves to dispense
with the necessity for furnaces of special construction and
to render it possible to use the burner gas produced by any
furnaces of ordinary construction, including the gas from
roasting zinc blendes or pyrrhotites, or, in fact, any metal-
lurgical gas.

This converter is based on the fact that in comparison to
the amount of pyrites desulphurized to produce the sulphur
dioxide, a much smaller quantity of ferric oxide than the
ore produces will suffice to oxidize the sulphur dioxide pro-
duced to sulphur trioxide; so that the heat produced by
roasting the larger part of the ore can be avoided or regu-
lated by roasting the ore in ordinary burners and conducting
the burner gas, at a comparatively low temperature, to a
converter in which only enough pyrites is burned to main-
tain the proper temperature and at the same time produce
sufficient ferric oxide for the contact substance.

65. The Frasch converter, shown in Fig. 10, consists of
a steel cylinder ^, similar to a cupola furnace, lined with
firebrick. Pyrites are charged into the converter through
the hopper b by means of the bell c. The ferric oxide is
discharged at the bottom into the double-valve hopper d^ so
as to prevent the admission of air during discharging.
This converter is on the down-draft principle. Air is
admitted through the pipes e and the products of combus-
tion carried away through the pipes/. When the furnace
is lighted and supplied with iron pyrites, a bed of burned

60 SULPHURIC ACID § 27

pyrites (ferric oxide) is formed, in which there will be vari-
ous zones of temperature from the upper to the bottom
layer of its contents. These zones of temperature can be
largely governed by the quantity of pyrites charged to or
discharged from the furnace, but in any case a zone of fresh
ferric oxide of suitable temperature can be maintained in
the furnace at some point. Burner gas (containing sulphur
dioxide) from outside sources, whether ordinary pyrites
burners or metallurgical furnaces, are now admitted
through the pipe g, and in passing through the zone of
ferric oxide of suitable temperature, the sulphur dioxide is
converted into the trioxide.

66, Purification of Burner Gas. — The burner gas, as
it comes from the desulphurizing plant, always contains
some, and often many, impurities. Of these, flue dust,
hydrofluoric acid, arsenic, and selenium have a most detri-
mental effect upon the contact mass, partly chemical but
principally mechanical, as they tend to glaze over and
destroy the porosity of the mass, thus rendering it inert. It
is further desirable to prevent the formation of dilute sul-
phuric acid, and its corrosive effect on the apparatus and
connections, by at once extracting the sulphur trioxide and
moisture contained in the burner gas.

This can be readily accomplished by first passing the gas
through a tower constructed in every respect as a Glover
tower, which is described later, except that it is packed with
smaller pieces of quartz. This tower acts as a scrubber and
collects most of the impurities, at the san^e time cooling
the gas to a point where it will more readily deposit the
impurities still remaining, in the next purifying apparatus.
The heat of the gas also concentrates such weak acid as is
formed by the sulphur trioxide and moisture contained in
the gas, together with such additional water as may be
found necessary to run down the tower. A necessary pro-
portion of this acid, when concentrated sufficiently (to
02'^ Baume), and separated from solid impurities by settle-
ment, may be used in the next apparatus to absorb the

§27 SULPHURIC ACID 51

moisture driven from the first scrubbing tower. After
absorbing this moisture in its dilute condition, it is again
run over the first tower and again concentrated, together
with the new acid formed in the first tower. The unused
increment, ultimately representing the daily quantity of
sulphur trioxide contained in the burner gas, is, if pure
enough, passed on to be further strengthened by the addi-
tion of sulphur trioxide in the main part of the contact
plant. If impure, it is sold or used for purposes for which
it may be suitable.

67. This first tower also serves another valuable pur-
pose, in that the heat from the burner gas concentrating
the dilute acid in the tower forms a considerable volume of
steam, which is intimately mixed with the burner gas pass-
ing through it. This admixture with steam prevents the
formation of volatile hydrogen compounds of the impurities,
especially of arsenic, phosphorus, or their compounds, which
would otherwise be formed by the action of the concentrated
sulphuric acid on the metal of the coolers and the impuri-
ties, and which could only be removed with difficulty.

After passing through this first tower, the gas is taken
through a long connection to the bottom of a second tower,
through which it ascends, meeting a flow of sulphuric acid
of at least 62° Baum6 (concentrated acid from the first
tower). This tower is constructed exactly like a Gay-
Lussac tower, which is also described later, except that it is
packed with very much smaller pieces of quartz or coke. In
this tower the gas is dried and deposits nearly all thti
remaining impurities.

The burner gas is now passed through a tower of the
same construction as the last, but which is dry (neither
water nor acid being used) and serves as a final drying filte?
and cooling apparatus.

68. other Methods of Piirlfyinjj: the Burner Gas.

The above description of the tower system of scrubbing the
gas sufficiently discloses the various purifying operations
necessary. Other apparatus merely accomplish the same

64 SULPHURIC ACID § 27

as may have been found most advantageous in each indi-
vidual plant, and with each special contact mass used, and,
of course, under any circumstances, between the tempera-
tures necessary to start the reaction and the dissociation
point of sulphur trioxide. The gas issuing from the contact
oven is now a mixture of sulphur trioxide, nitrogen, and
excess of oxygen, and, with a properly working process,
very small quantities of sulphur dioxide; and nothing
remains but to absorb or dissolve the sulphur trioxide in
water, allowing the inert nitrogen and oxygen to pass from
the apparatus into the atmosphere.

73. This is usually done on the principle of the reflux
cooler ; that is, the gas is passed through or over and in the
opposite direction to that of a stream of water or weak acid.
Consequently, the strongest gas meets the strongest acid
and the weakest gas meets the weakest acid, which more
readily absorbs it. As the absorbing apparatus is generally
of wrought iron, it is usual to start the process with acid
not weaker than 60° Baum6.

The combination of sulphur trioxide and water is also
exothermic.

SO^ + 1/^0 = H^SO, + 39.2 Cal (Thomsen)

73. Diagram of Contact I*roces8. — In Fig. 11 is shown
a diagram of the apparatus used in a sulphuric-acid plant
employing the contact process. The course of the various
materials and products is indicated by the arrows. A is 2i
bench of pyrites burners. The burner gas passes through
the flue a^ to the first cleaning tower B. Weak sulphuric acid
is constantly flowing down this tower, becoming concentrated
by the hot burner gas and absorption of the sulphur trioxide
contained in the burner gas, and finally flows out at the bot-
tom into the cooler (7 at a strength of from 62° to 64°
Baume. From the cooler C, the strong acid passes to the
tank D and is delivered by the pump D^ to the storage
tank T, or to the tank F over the second cleaning tower E,
A constant stream of strong sulphuric acid from the tank F

SULPHURIC ACIIJ

is kept flowing down this
tower. In tliis tower, the
burner gas coming from
the top of /; is further
cleaned and then passes to
the drying tower /; the
circulation of the gases
through the train of ap-
paratus is maintained by
the fan/. Hefore entering
the contact ovens, the
mixed gases are reheated
to the proper temperature
for the combination of the
sulphur dioxide and oxygen
in the reheater K.

74. The contact oven /,

" consists of cast-iron rings

£ with perforated shelves, or

diaphragms, upon which is

placed the contact mass.

The sulphur trioxtde
formed in the contact oven
now passes through the
absorption cylinders J/,,
J/„ J/„ .1/.. These are -
cylindrical iron tanks con-
nected in such a way that
the gas passes from end to
end, meeting the weak
acid flowing in the opposite
direction. Both the gas
and the acid in J/, are
richest in sulphur trtoxide,
while in ,1/', the gas and
acid are weak, and such
weak acid absorbs sulphur

56 SULPHURIC ACID § 27

trioxide most readily. The strong acid, which is ready for
the market as it comes from J/,, is collected in the tank Q
and is delivered by the pump (2, to the storage tank R.

The gases coming from the last absorption tank M^ con-
tains still a small amount of unabsorbed sulphur trioxide.
In order to recover this, the gases are passed through the
tower iV, which is supplied with weak acid from the tank -P,
which absorbs the last traces of sulphur trioxide. The
nitrogen and oxygen remaining pass into the air through
the pipe o. The tank car 5 receives acid for shipment from
the storage tank R,

SULPHURIC ACID

(PART 2)

THE CHAMBER PROCESS

INTRODUCTION

1. We have seen fhat the oxidation of sulphur, under
ordinary conditions, produces so much heat as to render
the existence of the trioxide possible only to a limited
extent, except in the presence of a third material possessing
so-called *' contact" properties, such as pyrites, cinders,
spongy platinum, cupric sulphate, etc. Also, that some of
these so-called contact substances, while producing a chem-
ical reaction, remain themselves in the end unchanged, what-
ever intermediate reactions they may or may not have
taken part in. In some of the contact phenomena, such
intermediate reactions can be traced, or, at any rate, such
is the only way of accounting for them. In the case of
contact phenomena connected with the complete oxidation
of sulphur into the trioxide, it is apparently possible that
electrical action is set up, which permits the formation of
the trioxide either by converting the excessive heat into
another form of energy, or which renders the oxygen, free
or combined with the sulphur dioxide, more active. In any
case, the contact substance in the final result appears to
suffer no chemical, change or deterioration, but only the
inevitable mechanical loss in handling.

§28

For notice of copyright, see page immediately following the title pagQ

2 SULPHURIC ACID § 28

2. It will now be seen that the chamber process is in
nature a contact process, inasmuch as a definite chemical
reaction between large volumes of sulphur dioxide, oxygen,
and water is induced by a small quantity of nitrous oxide
iV,(7„ which is recovered unchanged save for mechanical
loss; and yet without which the reaction would not have
taken place. In this case of contact action, however, the
intermediate reactions have been studied and are fairly well
understood.

When using the nitrous oxides as contact substance, or
oxidizer, of sulphur dioxide, the presence of water is abso-
lutely necessary and, consequently, only a hydrate or
solution of sulphur trioxide can be formed. If water were
not present, sulphur trioxide would be formed, but it would
combine with the nitrous acid to form nitrososulphuric acid,
or chamber crystals, {//0){iyO^)SO^. Water dissolves
these crystals, forming sulphuric acid and releasing the
oxides of nitrogen. Furthermore, water must be largely
in excess of the quantity required to produce the hydrate
H^SO^, as otherwise the oxides of nitrogen would be
absorbed and retained in the sulphuric acid; in fact, it
must be so much in excess as not to produce an acid
stronger than about 69 per cent, of the monohydrate (54° to
55° Baum6).

3. Reactions of the Chamber Process. — The follow-
ing explanation of the reactions that take place appears to
be the most rational and the one that coincides most closely
with the conditions of the actual chamber process.

(1) 5(9, + HNO^ -^0= {//0){iVO^)SO^

(2) 2{//0){N0,)S0, + Hfi = 2//,S0, + Nfi,

If, in the above reactions, sulphur dioxide, nitrous acid,
oxygen, and water be simply taken in definite quantity, then
a definite quantity of sulphuric hydrate and nitrous oxide will
be formed according to the above equations. The reaction
would end and the excess of sulphur dipxide, if any, would
pass on unchanged; but in the presence of excess of air

§ 28 SULPHURIC ACID 3

and water the nitrous oxide is converted into nitrous acid,
according to the following equation :

(3) iV,C7, + H^O = tHNL\

which again combines, according to equation (1) with the
sulphur dioxide so long as the latter is present in sufficient
quantity.

Or, in the presence of excess of oxygen (air) and water
(vapor or steam), sulphur dioxide, nitrous acid, and oxygen
form nitrososulphuric acid (chamber crystals). This is
immediately decomposed by water into sulphuric hydrate
and nitrous oxide Nfi^. The sulphuric hydrate con-
denses in the apparatus as a stable compound, while the
nitrous anhydride, with the water, forms nitrous acid, and
the above reactions are repeated until the sulphur dioxide is
practically all converted into sulphuric hydrate H^SO^,

4. In addition to the above principal reactions, another
set of reactions appears to take place in the Glov«r tower
and the first part of the first chamber, that is, where the
sulphur dioxide is largely in excess, and in which the nitroso-
sulphuric acid is partially decomposed by it'. ,

(4) 2{HO){2VO,)SO^ + SO^ + 2Hfi = d//^SC\ + 2NO

the oxide thus formed combining directly with the sulphur
dioxide, oxygen, and water to form nitrososulphuric acid.

(5) 2S0, + 2NO + dO + Hfi = 2{//0){N0,)S0,

which is converted into sulphuric hydrate and nitrous oxide
according to equation (2).

If the above reactions could be started with the exact
quantities of nitrous acid, sulphur dioxide, water, and oxy-
gen necessary, it is evident, to secure a continuous process,
all that would be necessary would be to secure a continuous
supply of the exact quantities of sulphur dioxide, oxygen,
and water, and return to the beginning of the process the
nitrous oxide accumulated at the end of the process by
simply supplying any mechanical loss common to all com-
mercial processes.

SULPHURIC ACID

§28

This is approximately what is done in the chamber
process. The nitrous oxide cannot, however, be returned
direct, as the oxygen, being supplied as air, carries with it
a very large proportion of inert nitrogen, which must be
gotten rid of. It becomes necessary, therefore, to separate

Fio. 1

the nitrous oxide from the inert nitrogen in such a way
that the N^O^ can again be made available and the inert
nitrogen wasted into the atmosphere.

Advantage is taken of the power of the stronger solu-
tions of sulphur trioxide from 60° to ()<r Baume, to absorb
and retain the nitrous oxide in fairly stable solution.

§ 28 SULPHURIC ACID 6

(6) 2/7.5(9, + N^O, = 2(//0){N0,)S0, + I/,0

In other words, nitrososulphuric acid is formed. When dis-
solved in a large excess of the sulphuric-acid solution, the
product is termed iiitroiis vitriol. The nitrous anhydride so
absorbed can be set free, however, on dilution of the acid
and especially in the presence of sulphur dioxide. When
this nitrous vitriol is diluted, in the presence of sulphur
dioxide at the beginning of the process, so as to set free the
nitrous anhydride and complete the cycle, the reaction is
represented by equation (4) above given. The diagram in
Fig. 1 shows the chemical reactions that take place during a
complete cycle. To read it, begin at the center and follow
the direction of the arrows.

APPARATUS EMPLOYED IN THE CILIMBER

PROCESS

5. In the manufacture of sulphuric acid by the so-called
chamber process, the first essential piece of apparatus is a
sulphur or pyrites burner provided with some means of
nitrating the burner gas. Any of the burners previously
described may be used.

6. Nitrating Oven. — Fig. 2 (a) and {b) shows an attach-
ment to the burners by which nitrating by potting may be
accomplished. Fig. 2 {a) is a horizontal section through
the niter pots d^ and Fig. 2 (b) is a vertical longitudinal
section through one of these niter pots. The extreme end
of a bench of lump pyrites burners is shown at a. The
flues b from the burners enlarge into the niter ovens r, in
which are placed the cast-iron niter pots, or **pigs," d.
The cast-iron dishes e underneath the niter pots catch any
acid material boiling over from the pots and prevent its
penetrating the brickwork of the furnace. A cast-iron
hopper, or funnel, f provides for the introduction of niter
and- sulphuric acid into the niter pots, the acid being stored
in the tank ^ and conducted by a lead pipe and cock to the

§ 28 SULPHURIC ACID 7

hopper. The common flue and dust chamber // leads to the
cast-iron flue k^ through which the gas is carried to the Glover
tower.

When the burners are in operation, the pots d are supplied
with niter and a regulated amount of sulphuric acid added.
The fumes of nitric acid thus formed mix with the hot
burner gas and pass to the Glover tower. The sodium sul-
phate formed in the pots is removed through the cast-iron
neck /, which is usually kept closed with a wooden plug, into
the cast-iron dishes y. When cold and solid, it is broken up
and removed.

7. This method of nitrating by "potting" is by no
means satisfactory, because it adds another element of
periodic irregularity to what should be a continuous proc-
ess, and because, unless in the hands of careful and skilled
workmen, it is a wasteful and a dirty process. It is also
difficult in this way to supply the chambers with nitrous
oxide just in the quantity and at the time when it is
most wanted — that is, when something in the process is
going wrong. Sometimes, also, on account of faulty con-
struction, there is insufficient heat to decompose the niter
rapidly enough or else the heat is too great and too direct
and the sulphuric acid is evaporated before it has reacted
completely with the sodium nitrate.

Wherever, therefore, the size of the plant justifies the
manufacture of nitric acid on a small scale or where it is in
any way possible, nitration should be secured by the use of
nitric acid run into the Glover tower with the nitrous vitriol.
This is accomplished by means of a small glass siphon from
the nitric-acid tank or carboy, fitted with a glass cock and
discharging from the cock into a glass funnel with a bent
neck, so as to form a seal or lute and fixed into the center
of the top lead of the tower. It does not matter how weak
or impure the nitric acid may be for this purpose; indeed,
in some works, the spent acid from the manufacture of
nitroglycerin is used, as the acid is almost instantly decom-
posed upon entering the tower.

8 SULPHURIC ACID § 28

8. Glover Tower. — The apparatus in which the sulphur
dioxide, oxygen, and nitrogen of the burner gas are mixed
with the nitrous oxide N^O^, derived from the nitrous vitriol
used in this stage of the process, water vapor and the
nitrous fumes from the nitrating ovens, which after the
process is once under way is only sufficient to make up for
the mechanical loss, is known as the Olover tower. In
this tower, the gases and vapors are not only thoroughly
mixed, but the dilute sulphuric acid constantly flowing down
is both denitrated and concentrated by the hot gases, ren-
dering it strong enough to be again used for absorbing N^O^
at the end of the process.

9. The heat of combustion of the sulphur to SO^ in
the furnaces is usually more than sufficient to concentrate
the whole of the make of chamber acid if entirely util-
ized to 66° Baum6, or to 93.5-per-cent. H^SO^. In a well-
constructed plant, that is, where the heat is fairly well
utilized- the Glover tower will concentrate from one and one-
half times to twice the entire make of chamber acid to 60°
or 62° Baum6 (62- to 80-per-cent. H^SO^, or, in other words,
this quantity of chamber acid can be used to dilute the
nitrous vitriol and will leave the Glover tower at 60° Baume
or over. Of course, if it is not desired to keep this amount
of acid in circulation between the Glover and Gay-Lussac
towers, the nitrous vitriol may be diluted in whole or in
part with water.

The temperature of the burner gas entering the Glover
tower will vary, of course, with the construction and length
of connections, but will average probably about 550° C.
The greatest possible temperature produced by the com-
bustion of sulphur will, of course, vary with the nature of
the raw material. Mendeleeff estimates the highest possible
temperature of actual sulphur burning in air to be 1,974° C.
and in oxygen 7,258° C.

10. The construction of the Glover tower is clearly
shown in Fig. 3. It consists of a circular brick-lined tower e

g 88 SULPHURIC ACID 9

covered with a lead sheathing / and lead pan o at the bot-
tom, and is filled to near the exit pipe ^with a packing /

consisting of broken quartz, the pieces being large at the
bottom, but decrease in size towards the top. This packing

SULPHURIC ACID

§28

§ 28 SULPHURIC ACID 11

rests upon the grill tiles d, which are supported by the
walls b. The tank // contains dilute or chamber acid, which
flows through the equalizer //, and the distributor //, over
the top of the packing. On the other side is a similar
arrangement /, /,, and /, for the distribution of nitrous vit-
riol, which is strong sulphuric acid coming from the Gay-
Lussac tower, described later, and heavily charged with
nitrous oxide Nfi^^ this Nfi^ being set free on dilution
of the vitriol in this tower.

The burner gas enters the tower at the bottom by means
of the pipe ^, which is surrounded next the tower by the
cast-iron cooling ring ^, which prevents the heat from injur-
ing the lead sheathing next the pipe. The gas is distributed
through the gas spaces c and passes through the grill
tiling up through the packing, coming in intimate contact
with the dilute acids from above, which are giving up N^O^^
and become mixed with the latter and also with steam
formed by the hot burner gas on the dilute acid. This
mixture of burner gas, nitrous oxide, and steam passes on
through the pipe g into the first lead chamber.

As previously stated, the rapid evaporation of the moisture
concentrates the down-flowing acid considerably. The deni-
trated and concentrated acid having a strength of from 60°
to 62° Baume is drawn off at j\ the lead-covered cast-iron
plate, or dish, n catches the acid or other leakage. The exit
pipe k is for use when the tower is washed by flooding with
acid in too large quantity to pass through/. The tower is sup-
' ported on the foundation walls /and the I beams ;//.

In dimensions, the Glover tower will average about 24 feet
in height and 12 feet in diameter. The construction is
necessarily heavy, in order that it may withstand the high
temperatures.

11# Ijead Chambers. — The thoroughly mixed gases
from the Glover tower containing nitrous oxide Nfi^ and
water vapor are allowed to pass to the chambers in which
the oxidation of the sulphur dioxide to trioxide and the for-
mation of sulphuric hydrates takes place. These chambers

SULPHURIC ACID

§28

Mv^

r^^Vv^

J ,8

!//>;■ ^jy'.^-r Saba..

/■

/^^

.Vwi

§ 28 SULPHURIC ACID 13

are usually three in number, of greatly varying dimen-
sions, but average between 50 and 100 feet long by from
20 to 30 feet wide and 20 to 30 feet high. They are con-
nected together in series, the communication between them
being -comparatively small. The construction of the cham-
bers is shown in Figs. 4, 5, 6, and 7.

Fig. 4 is a side elevation, showing the method of framing.
The chamber building is built on posts n' upon which are the
corbels k' supporting the stringers /'. The joists m' are laid
on these stringers, and upon these are laid the sills a of the
chambers. The posts b and the intermediate uprights c are
erected upon the sills and stiffened by the braces d. The
crown tree e surmounts the posts and intermediates, and on
this the top joists g are laid. The floor of the chamber is
covered with sheet lead, so as to form a pan whose edge is
shown at i. The edge of the lead curtain forming the inside
lining of the sides is shown aty. The end wall of the cham-
ber building is shown at o',

13. In Fig. 5 (a), {b), (r), and (d) is shown the method of
attaching and supporting the lead lining. Fig. 5 {a) shows
the method of cutting the lead straps for supporting the lead
lining. Fig. 5 {b) shows the top joists g with the lead
straps ;/ attached, the lower ends of the straps being burned
to the top lead ///. Fig. 5 {c) is a plan of the top, showing
the method of fitting the lead lining into the corners. The
top lead /;/ is supported from the top joists g. The crown
tree is at e. A long horizontal strap /t is nailed to the crown
tree and supports the side lead at the top where it is
attached to the top lead. The attachment of the top and
side leads is best shown in Fig. 5 (^), which is self-explana-
tory.

In Fig. 6 {a), (^), and (^), further details of the attach-
ment of the side and pan lead are shown. Fig. 6 {a) shows
a horizontal section through the posts b and the uprights c
at a corner, showing the attachment of the side straps / to
both posts and lead. Fig. 6 {b) is a side elevation towards the
bottom of the chamber, showing the method of attaching

SULPHURIC ACID

§28

(f)

mj\

■^

^JMJ ]IM LliaHI-^-^

(r> /i

Pig. 6

§28

SULPHURIC ACID

16 SULPHURIC ACID g 28

•

the sides of the lead pan /by rolling the top over the strip/.
Fig. 6 (c) is a vertical section through a side, showing the
relative positions of the side lead/ to the pan /.

This pan is kept about two-thirds full of acid and at all
times the curtains or sides should dip at least 2 inches into
the acid. When it is desired to draw acid from the cham-
bers, it is done by means of the arrangement shown in section
in Fig. 7. A pipe o is burned into the bottom of the cham-
ber; the entrance to this pipe is protected from the wash of
the flowing acid and a stratum of cool acid is kept on the
bottom by means of a loose lead ring /, which may be
removed when it is necessary to entirely empty the chamber.
The pipes from two or more adjacent chambers meet in the
cylindrical lead boot g. This boot is provided with a lead
plug r or valve and seat communicating with a pipe s lead-
ing from the chambers to a tank or wherever it can flow by
gravity. The entrances / to the boot from the chambers
can also be plugged, so that acid can be drawn from either
chamber or both, and the level in the two chambers can be
regulated as desired.

SURFACE CONDENSERS

13« Immediately on the entering of the gas into the
chambers, the formation of sulphuric acid commences.
This acid is formed as a very fine mist. This mist gradu-
ally and slowly settles on the sides and bottoms of the
chambers. As the gas leaves the first chamber it is very
advantageous to condense this mist of already formed acid
that it contains, so as to leave the gas free to enter into
renewed activity upon entering the second chamber. The
same thing may be said of the gas leaving the second cham-
ber and entering the third chamber. Many proposals have
been made to secure condensation at these points.

14, Iiunge Condenser. — Lunge has introduced what he
calls plate columns for this purpose, consisting of a lead
tower, or column, fitted with flat, perforated, earthenware

SULPHURIC ACID

18 SULPHURIC ACID § 28

plates in layers one above the other and about % inches
apart. A stream of chamber acid is run over the plates.
The. perforations are so arranged that the acid in dropping
through the perforations of one plate splashes upon the solid
part of the plate below it and is thus broken into spray, up-
on meeting which the gas is cooled and deposits its mist of
contained acid. This apparatus, therefore, may be con-
sidered as a type of spray condenser, similar to the well-
known form used in steam engineering.

Fig. 8 shows the Lunge type of spray condenser. The
gas is admitted at a into the lead-lined box ^, whence it
passes through the perforations in the plates c, c, meeting the
stream of acid supplied by the distributors d and lutes c.
This acid, together with the condensed mist contained in
the gas, is collected in the pan /and either run back into one
of the chambers or conveyed by lead pipe to storage. The
gas passes on to the vent chamber / and through the col-
lecting pipes ^, ^„ ^„ ^„ and the main pipe A.

15. Gilchrist Condenser. — The Gilchrist pipe col-
umns consist of an oblong tower, or column, of lead
pierced in its smaller diameter by a series of lead pipes open
to the air at each end. The lead column is surrounded by a
wooden breaching and flue in such a way as to cause a cur-
rent of air through these pipes, thus tending to keep them
cool. The gas passing through this column is cooled by
contact with these pipes and the acid mist is condensed on
them. This apparatus may therefore be considered a type
of air-cooled surface condenser-
Fig. 9 shows the Gilchrist air-cooled surface condenser.
The gas is admitted at a into the lead box i. This box is
pierced by numerous lead pipes c, r, ^, open at both ends.
The acid mist contained in the gas is condensed on these
pipes and the comparatively cool surfaces of the lead box
and runs to the bottom of the box ^, whence it is carried to
a chamber, or storage, by a pipe d. The gas then passes
through the collecting pipe e to the vent chamber. The
lead box is surrounded by a wooden breaching, so that the

SULPHURIC ACID

air entering^ from below is drawn through the lead pipes
c, c, c into the breaching g' and thence to the draft pipe J,
thus tending to keep the apparatus cool.

16. The Faldlng Condenser. — The Palding surface
condenser consists of a series of lead pipes surrounded by
water as the cooling medium. They are arranged in such
a way as to secure a maximum efficiency with a minimum
use of water. This apparatus may therefore be considered
as a type of water-cooled surface condenser.

Fig. 10 shows the Falding water-cooled surface con-
denser. In this condenser, the entering gas is broken up
into a number of small streams through lead pipes a, a, a, a.
These pipes dip almost to the bottom of a series of water-
cooled lead pipes d,d, of larger diameter, with closed bot-
toms. The annular space between these pipes contains a
strip of lead, which forces the gas to return in a spiral
through acid to the top of the annular space, whence it

SULPHURIC ACID

§28

■49/y/jj/^^

§ 28 SULPHURIC ACID 21

passes through pipes r, c into the next chamber or into a
header or manifold and thence into the next chamber. The
condensed acid mist runs from the apparatus at b.

17. Other Condensers. — Many manufacturers use sim-
ple lead towers filled with quartz, brick, or special earthen-
ware shapes. These towers do not take sufficient account
of the necessity for cooling, and while they are efficient to a
certain extent, they are not sufficiently so when their cost
relative to an equal amount of chamber spaces is taken into
consideration.

If all operations have been properly conducted, the gases
coming from the last lead chamber are practically free from
sulphur dioxide, and consist of inert nitrogen, the excess of
oxygen, and nitrous oxide N^O^, This latter gas, if freed
from the other two gases, may be used over again as
an oxidizer for more sulphur dioxide. This separation
depends on the fact that nitrous oxide N^O^ is readily
absorbed by concentrated sulphuric acid forming the
so-called nitrous vitriol, while the other useless gases are
unabsorbed. The apparatus in which this absorption takes
place is called the Gay-Lussac tower,

18, Gay-IiTissac Tower. — This piece of apparatus is in
construction very similar to the Glover tower, but dif-
fers from it in that it is of somewhat lighter build. Its
height is greater, the average height being about 50 feet, and
its diameter is somewhat less, being about 8 to 10 feet.

The details of the Gay-Lussac tower are shown in Fig. 11.
The brick walls e are of light weight and are covered with a
lead sheathing/. Under the brick bottom is the lead pan o
resting in the lead-covered cast-iron dish n. The tower is
supported on the I beams ;// by the foundation walls /.

The filling /is of broken quartz, coarse at the bottom but
becoming finer at the top, as in the Glover tower. The
tank h contains strong, 62° Baume, sulphuric acid, which
flows through the- equalizer / and the distributors s over the
top of the packing.

SULPHURIC ACIU § is

§ 28 SULPHURIC ACID 28

During operation, the mixed gases from the chambers
enter at the bottom through the pipe a, pass through
the gas spaces c in the supporting waJl ^, and up through
the grill d into the packing material. As the gases ascend,
they come in contact with the descending concentrated
sulphuric acid, which absorbs the N^O^, The unabsorbed
gases pass through the pipe g into the air or, more com-
monly, into a second Gay-Lussac tower, which absorbs any
Nfi^ that may have escaped absorption in the first tower.
The nitrous vitriol is drawn oflf at the bottom of the tower
aty. The exit k is for flushing purposes.

The nitrous vitriol coming from the Gay-Lussac tower is
pumped to the tank over the Glover tower and is used in
the Glover tower, where it gives up its iV,C7„ which again
passes through the system.

19. Diagrram of Chamber Process, — The disposition
of the various pieces of apparatus already described and the
cause of the various materials and products is indicated in
the diagram shown in Fig. 12. Reference to this diagram
will enable one to keep a general idea of a plant in mind and
better understand the process as the details are dis-
cussed.

In the figure, /I is a bench of pyrites burners, niter oven,
etc. The burner gas is conducted through the pipe d to
the Glover tower E, where it meets the dilute acids and
oxides of nitrogen. The fan y carries the gases through the
pipe t to the first chamber K, where oxidation of the sulphur
dioxide takes place, thence to the second and third cham-
bers iWand Ny through the flues /, and /, and surface con-
densers L and Z,. The acid drained from the bottom of each
chamber and the condensers is collected in the tank R^,

The pump 5, of one of the styles shown in Figs. 13 and 14
delivers this acid to the tank H^, over the Glover tower, or
to the storage tank Uy whence it goes to the tank car V. The
strong acid coming from the Glover tower is collected in
tanks Q and R^ and is delivered by the pump S, to the
tank //, over the second Gay-Lussac tower P and to the

<9V

I /

■S

StHr*

c«

§ 28 SULPHURIC ACID 25

storage tank U^, The gases from the last chamber TV are
conducted through the pipe /, to the first Gay-Lussac
tower O and thence to the second Gay-Lussac tower /*,
their flow being maintained by the fan^,. The exhausted
gases pass to the atmosphere at /. The nitrous vitriol from
the first Gay-Lussac tower is collected in the tank R^ and
is delivered by the pump i^ to the tank // over the Glover
tower. The nitrous vitriol from the second Gay-Lussac
tower, containing but little Nfi^^ is collected in the tank R^
and' is delivered by the pump S^ to the tank //, over the first
Gay-Lussac tower. In different works, this scheme varies
somewhat in detail, but not in its essential points.

20, Acid Pumps. — In both the catalytic and chamber
processes, it is necessary to transfer large volumes of acid
from one part of the works to another. This is done by
means of pumps of peculiar construction, some of which are
designed to act automatically, so as to give a continuous
flow of acid. Two styles of pumps, the Kcstncr automatic
and Monteju's acid egg, are here described.

21, Kestner Automatic Pump. — This apparatus,
shown in Fig. 13, is automatic and works continuously; it is
constructed of cast iron for strong acid, but is lead lined for
weak acids. It is operated by compressed air. The acid
chamber is connected by the vertical pipe b with the valve
box c, which must be placed higher than the tank supplying
the apparatus, so that in no case acid can rise within a foot
or two of it. Acid is admitted from the supply to a
by means of the pipe d and check-valve e. The float f con-
nected with the counterbalanced compressed-air valve g
by means of the rigid rod h running inside the vertical pipe b
and stuflingbox f, is raised by the inflowing acid until it
opens the compressed-air valve g. The compressed air from
the pipe /communicating with taty* flows through the pipe b
into the acid chamber a, driving the acid up through the
pipe >& to a receiving tank ; for instance, on top of a tower.
As soon as chamber a is empty the float falls, closing the air
valve, and the operation is repeated. The air valve and

SULPHURIC ACID

float are so balanced that the total movement of the rod
does not exceed ^ inch. The great advantage of this appa-
ratus is that it insures a steady flow of acid (which can be
accurately controlled) over the towers.

^'4. M(>iitejii*!s Pump With Avid Egg. — This pumping

arrangement is illustrated in Fig. 14. The tank .^ contain-
ing the acid communicates at f with the receptacle or
"egg " G by means of the pipe 6, the flow being controlled

28 SULPHURIC ACID § 28

by the globe valve c. The plug valve V is merely auxiliary,
and should not be relied on, as it can only with difficulty be
made to withstand the back pressure. The check-valve d
is used under ordinary circumstances. This valve permits
the flow of acid into the ^%% until the acid rises to the level
of the valve, which, when the compressed air is let into the
egg» immediately seats itself and prevents the air from for-
cing the acid back into tank A,

Compressed air is admitted to the t,%% by means of the
pipe /' and the valve /. The pipe h controlled by the
valve / delivers the acid from the ^%% to the splash box y of
the distributing tank P, When air is admitted to the ^%%y
as it cannot pass valves d and r, and valve i being open, it
forces the acid to a height of from 50 to 100 feet through It
into the splash box j\ which is a lead-lined box with two
openings, through the lower of which the acid escapes into/',
an open part of the tank, and thence through the exit ni into
a receiving tank on top of the towers and an upper opening
of large area, whereupon the air escapes into the atmos-
phere without splashing the acid over things. The exit n
from y into another receiving tank is provided in case the
^%% is used for pumping two kinds of acid, the plug being
simply moved from n to ;;/ and a branch connection to a
second supply tank being inserted at ^, the flow of acid from
either supply tank into the ^%% being then controlled by
plugs b\

OPERATION OF THE CHAMBER PROCESS

23, If the reactions involved in the chamber process
have been understood, the importance of extreme regularity
both aG to volume and composition, of the supply of the
substances entering into these reactions will appear obvious.
For, although the process involving these reactions is a con-
tinuous one, and in fact more especially on this account,
if loss is to be avoided and success attained, the supply of
the necessary ingredients must be as exact as if the proc-
ess were an isolated reaction involving the complete union

§ 28 SULPHURIC ACID 29

of carefully weighed proportions. The materials in ques-
tion are: (I) A constant stream of burner gas of uniform
volume and percentage of sulphur oxides and free oxygen.
(2) A uniform supply of finely divided water or water vapor
of constant tension. (3) A uniform supply or circulation
of nitrous vitriol containing a constant percentage of nitrous
oxide N^O^, (4) A uniform supply of nitric oxide or
acid for making good the oxides of nitrogen lost in the
process (mechanically or otherwise).

It is only by careful watchfulness, honest work, and proper
management, together with a rationally constructed plant,
that a near approximation can be made to the requirements
as to absolute uniformity called for. When, however, such
Approximation is reached, the difficulties of the chamber
process disappear and the operation will proceed month
after month with little, if any, variation, and with uniform
results.

24, Conditions in the Glover Tower. — The burner
gas, having an average temperature of about 550° C, in
passing from below through the Glover tower meets a finely
divided stream of nitrous vitriol 2f/^S0^ + N^O^ greatly
diluted with chamber acid or with water, or both, and often
carrying with it nitric acid, sufficient to supply the loss
inevitable in the process amounting from 1.5 to 3 per
cent. (The consumption of oxides of nitrogen is always
given in terms of percentages of sodium nitrate NaNO^ cal-
culated on the available sulphur burned.) This stream of
mixed acids enters the top of the tower at from 40° to 50°
Baum6, according to the degree of concentration and deni-
tration required and the concentrating efficiency of the
tower. The hot, moist, sulphurous gas drives off the nitro-
gen oxides in the upper part of the tower, and as it
descends to the lower and hot zone, the water is expelled
from the dilute acid as steam. The acid is thus concen-
trated to from 60° to 62° Baume, or in special cases to
64° Baum6, or even to 66° Baum6 and flows from the
tower, while a stream of gas containing a mixture of oxides

30 SULPHURIC ACID § 28

of sulphur and nitrogen, steam, oxygen, and nitrogen,
passes over to the first chambers.

36. Conditions In the Chanibers. — The gas thus enter-
ing the first chamber contains all the elements necessary
for the production of the hydrate or solution of sulphur tri-
oxide and in a condition of maximum activity. At this
point, the percentage of sulphur oxides is greatest, the free
oxygen is in greatest excess, and the oxides of nitrogen NO
and Nfi^ are such as possess the most f>owerfully oxidizing
effect. The temperature of the gas (80° to 100° C.) is also
conducive to an active reaction.- Therefore, it is at this
zone of reaction that one would naturally look for a large
make of acid, and such is actually the case, for between the
Glover tower and the first forty feet of the first chamber,
with all the elements and conditions of the process at their
best, from 60 to 80 per cent, of the whole acid is made.

36. In a properly constructed plant, that is, a plant con-
sisting of rightly proportioned Glover tower, chambers,
and Gay-Lussac towers, a sufficient quantity of nitrogen
oxides should be supplied to the gas by means of the Glover
tower to raise the temperature of the reaction (as shown by
the thermometers penetrating the sides of the chambers,
say at a distance of 25 feet from the end that is nearest the
Glover tower) to from 95° to 100° C. This, of course, does
not apply to the oxides of nitrogen supplied to the system
to replace the mechanical loss, but to the nitrogen oxides
recovered at the end of the process and gradually accumu-
lated as nitrous vitriol (nitrososulphuric acid dissolved in a
large excess of 60° to 62° Baume sulphuric hydrate or solu-
tion) and which is run over the Glover tower in dilute form
to again utilize its contained oxides of nitrogen. The
oxides of nitrogen so stored may be termed niter in circula-
tion^ and it is evident that, according to the quantity of
this nitrous vitriol of uniform percentage contents of nitro-
gen oxides accumulated, put into circulation at the Glover
and recovered at the Gay-Lussac towers, so will be the ratio

§ 28 SULPHURIC ACID 31

of active nitrogen oxides to the sulphur oxides at this crit-
ical initial point; i.e., the Glover tower and first part of
the first chamber.

27. Provided always that the towers are properly pro-
portioned to fulfil their functions of denitration and absorp-
tion (or recovery), it is desirable to accumulate and put into
and keep in circulation about 20 per cent, of niter (by niter
is meant oxides of nitrogen calculated as nitrate of soda
NaNO^ on the available sulphur burned). This will secure
an active process at the beginning and a rapid oxidation of
the gradually lessening percentage of oxides of sulphur
after the first active zone has been passed, owing to the
large excess of active oxides of nitrogen in the chamber
gas, and, consequently, a rapid change of these oxides of
nitrogen to nitrous oxide iV,0„ in which form it is capa-
ble of being at once absorbed in the Gay-Lussac tower.
This will, on the other hand, prevent the process becoming
sluggish and slow, with the consequent danger of sulphur
dioxide escaping into the Gay-Lussac tower unoxidized,
where it will decompose and so prevent the complete absorp-
tion of the nitrous oxide by the sulphuric acid, which takes
place according to the following equations:

2{HO){NO;)SO^ + S0^'+ %HJ0 = 3//,5(9, + ^NO

The oxide NO will not be absorbed, but passes with
the inert nitrogen into the atmosphere. It will also avoid
(by at once absorbing from the process) the danger of the
Nfi^ being changed to NO^ or even to nitric acid HNO^,
when in the first case it would be lost as stated above, or
in the second case it would not only be lost but would
rapidly destroy the lead of the apparatus and contaminate
the acid made.

38, After the first 40 or 50 feet of travel of the gas in the
first chamber, the temperature indicated by the side ther.
mometers will rapidly diminish. This would naturally be
expected as the reactions become less intense, on account of

32 SULPHURIC ACID §28

the lesser proportion of sulphur dioxide contained in the
gas, and also its greater diffusion in the chamber and its
saturation with a mist of already formed sulphuric hydrate.
The length of the active zone, of course, varies according to
the volume of burner gas passed into a chamber of any given
size, and also to the intensity of the first reactions, depend-
ing on the proportion of nitrous vitriol kept in circulation ;
but sooner or later, and generally within the first 60 feet,
the reactions, as indicated by the thermometers, will become
sluggish and will so continue until the gases have been thor-
oughly mixed and the various elements brought into more
intimate contact by passing them through a pipe connection
and in their mixed condition allowing them to again expand
in a second lead chamber. For this reason, it is now usual
in the United States to limit the length of the first chamber
to from 50 to 75 feet.

39, Where a positive method of controlling the currents
of a gas (such as the use of fans, etc. ) exists, it is preferable,
in the case of large volumes of burner gas being handled, to
divide the gas between two or more first chambers of lim-
ited length, so as to. secure a large zone of great activity
rather than an extended zone of rapidly diminishing activity
or sluggish reaction.

The condition of the gases at the end of the first cham-
ber, or after the zone of great activity, is such as to call not
only for a thorough mixing but also for a cooling and a con-
densing of the mist of acid already formed. Radiation of
heat from the surface of the chambers, while very consid-
erable, is not sufficient by itself to conduct away the heat of
the active zone so as to secure the best results. The tow-
ers, surface-, air-, and water-cooled condensers and plate
columns employed have already been described. These
apparatus, by bringing the gases again into intimate con-
tact, also undoubtedly start the reactions into renewed
activity.

30, The second chamber in a properly proportioned set
and with sufficient nitrous vitriol in circulation (in other

§ 28 SULPHURIC ACID 33

words, with a sufficiently active process) will almost entirely
oxidize the remaining sulphur dioxide, so that with or with-
out further surface condensers between the second and the
third chamber, the oxidation will be completed at once on
entry into the third chamber, which then acts merely to dry
and cool the gas, now consisting of inert nitrogen, the excess
of oxygen, and nitrous oxide, and render it fit for absorption
in the Gay-Lussac towers. For cooling and drying the gas,
a long pipe connection between the last chamber and the Gay-
Lussac tower is of great advantage; it can, however, be
replaced by a surface condenser of any of the types pre-
viously mentioned.

In this description of the passage of the gas through the
sulphuric-acid plant, it must be remembered that while the
gas enters the chambers containing a large proportion of
water vapor derived from the concentration or evaporation
of the dilute acid supplied to the Glover tower, this water
is rapidly absorbed by the formation' of the sulphuric
hydrate and precipitated to the pans of the chambers.

More water, either as finely divided spray or as steam,
must be added. Steam is the usual medium employed,
either low-pressure steam (20 pounds per square inch) or
exhaust steam from a neighboring engine, or both.

31. Admission of Steam to the Chambers. — It is

well to have sufficient points of admission for the steam,
either on the top or sides of the chambers, each point being
supplied with an indicating valve, so that the steam may
ultimately be supplied just at such points and in such quan-
tities as experience may show to be the best in each individ-
ual case, and under varying conditions of conducting the
process. Just as it is with the burner gas and the supply of
nitrogen oxides, so must the flow of steam to the process be
in every respect uniform. To secure this, the steam pipes
must be well covered and trapped and the main line sup-
plying steam to the branches must be supplied with steam
gauges and an efficient reducing valve, which must be con-
stantly watched and kept in order. The arrangement of

SULPHURIC ACID

§28

3£

1£

(14

§ 28 SULPHURIC ACID 35

the steam-pipe connections is shown in Fig. 15. The main
supply pipe u is laid between the chambers, the vertical
pipe ^ extending from it to the top and having branches 7*
to the chambers right and left. The lead terminal pipes x
enter the chambers by means of the hydraulic lutes/, which
are ordinary water seals. At m is the top lead of the cham-
ber. The indicating valves w serve to regulate the flow of
steam to the chambers. If steam is admitted to the sides
of the chambers, the lead terminal pipes enter the side
leads or curtains through specially constructed stuffing-
boxes.

With uniformity in the supply of gas, nitrogen oxides,
and steam, and a draft subject to proper control once
started, the chamber process becomes continuous and simply
requires careful watching to maintain the regularity of the
conditions. A careless burner man, by admitting too much
air to the furnaces and thus reducing the percentage of sul-
phur dioxide in the burner gas, or a careless tower man in
sending an irregular flow of nitrous vitriol over his Glover,
will very quickly destroy the harmony of the reactions and
too quickly disarrange the process to such an extent that
first the supply of nitrogen oxides in circulation and then
the sulphur dioxide itself will be pouring out into the atmos-
phere and the process will have resolved itself into the
same, or almost the same, conditions, the acid maker has to
confront when ** starting up his chambers*' or, in other
words, at the beginning of everything.

32. StartiLngr the Chamber Process. — This part of
the operation requires the exercise of care and judgment,
and it will take from 24 hours, where fans are used, to three
or four times as long before the process is normal. The
Glover tower, with its massive packing, absorbs much heat,
and it will take considerable time for it to reach a tempera-
ture at which it will perform its double functions of denitra-
tion and concentration in a satisfactory manner, more espe-
cially as the acid that must be run into it from above has a
constant cooling effect. At the same time, the Gay-Lussac

36 ' SULPHURIC ACID § 28

towers become saturated with sulphur dioxide, which
prevents the proper absorption of the nitrous oxide, and
the formation, consequently, of a stock of nitrous vitriol
for the Glover tower. These difficulties, of course, are
exaggerated where no stock of 60** to 62** Baum6 acid or of
nitrous vitriol is on hand, and where the process has to be
started with a supply of chamber acid alone (or even of
water), as is generally the case in an isolated chamber
system.

When such is the case, the chamber pans must be filled
with sufficient acid of from 50° to 54"* Baum6 to form a
hydraulic lute with the curtains or side and end sheets of
the chamber lead.

A small quantity of acid must be run down the Glover
tower until the packing is thoroughly moistened, and the
Gay-Lussac towers should also be supplied with acid.
Whether nitrogen oxides are to be supplied by ** potting,"
or by the direct use of nitric acid on the Glover tower,
arrangements must be made that will enable an abnor-
mal amount to be used until such time as the towers
are working properly and the stock of nitrous vitriol for
circulation is secured. It will be advisable, a\. first,
to supply an amount of nitrogen oxides equal to at
least 8 or 10 per cent, of sodium nitrate, on the available
sulphur.

The burner gas is then turned into the Glover tower and
the chamber system. At first and until the Glover tower
is performing its functions properly, it will be necessary to
supply steam to the first part of the first chamber. This,
however, will have to be done with extreme caution, as too
great an excess of water is likely to cause the formation of
nitric acid HNO^^ which will cause the rapid deterioration
of the chamber lead.

33, As the Glover tower gets hotter it will concentrate
the limited amount of acid with which it is supplied, to
about 60° Baum6, and the quantity of acid can then be
gradually increased. This stronger acid is at once supplied

§ 28 SULPHURIC ACID 87

to the Gay-Lussac towers, which will then commence to
absorb a little nitrous oxide; with patience and watchful
care matters will gradually assume a normal condition.
A sufficient stock of nitrous vitriol having been accumulated,
and the steam admission, pumping arrangements, and the
flow of acid over the various towers regulated, Vie extra
niter supply will be reduced to a point where it is just
sufficient to supply the daily loss and maintain the circula-
ting supply of nitrous vitriol intact. The acid concentrated
by the Glover tower should test 62'' Baum^ at 60° F.
(66.4-per-cent. SO^), Such part of it as is intended to be
run over the Gay-Lussac towers should be run from the
Glover tower into a cooler and cooled as thoroughly as the
temperature of the cooling water will allow. It is then
pumped to the supply tank on the second Gay-Lussac tower,
where it meets with the gas just leaving the system and
poorest in N^O^. It will run from this tower containing vary-
ing percentages of nitrososulphuric acid, and is known as the
first, or weak nitrous vitriol. It is then pumped to the first
Gay-Lussac tower, or the tower nearest to the last chamber,
where it meets the gas strongest with A^, (9,. Sufficient acid
should be supplied to these towers to permit a nitrous vitriol
containing 2.5 to 3 per cent, of J^^O^ to run from this first
towei:. This second, or nitrous vitriol, proper, ' is then
passed to the stock tanks for nitrous vitriol, an exactly equal
amount, both in quantity and percentage of N^O^y being
taken from the stock tanks and pumped to the top of the
Glover tower and run down the tower together with a suffi-
cient stream of weak sulphuric acid to dilute it sufficiently
to secure denitration and also to secure its concentration in
the Glover tower to 62° Baum6.

34, All well-equipped plants are now being built
with two Gay-Lussac towers, both because in this way it is
possible to secure sufficient cubic capacity without undue
height or diameter, and because if, for any reason, the proc-
ess becomes irregular (**goes back ") and sulphur dioxide
gets into the first tower, decomposing the nitrous vitriol,

38 SULPHURIC ACID § 28 •

then the second tower will still absorb and to a consider-
able extent take up the work which the first tower is doing
badly, the first tower, in the meantime, assuming the func-
tions which should have been performed by the last chamber.
In this way, time is secured to find out just where the
trouble is and remedy it before much harm is done. If, •
however, the trouble is not found and remedied, the sulphur
dioxide will gradually get into the second tower and the
process will be *'lost," or in other words, with the excep-
tion that the Glover tower is hot, the acid maker will have
to proceed as in starting up the system.

35, It must be borne in mind, and too great emphasis
cannot be given to the statement, that when the chamber
process begins to go wrong, // is on account of a break in the
uniformity of the supply of the various elements. Either the
burner gas is richer or poorer in sulphur dioxide, the nitrous
vitriol is poorer in nitrous oxide on account of the acid
supplied by the Glover tower being weaker than 62"^ Baum6,
or too much or too little steam or higher or lower pressure
steam is being supplied. When such irregularity is noticed,
the acid maker must at once increase the flow of nitrous
vitriol from his stock over the Glover tower. He will then
immediately test his burner gas, nitrous vitriol, steam, etc.
until he finds where the irregularity is occurring. This
remedied in time, the process will rapidly become normal
again and the increased supply of nitrous vitriol may be
cut off gradually, in the meantime more 62° Baum6 acid
being run over the Gay-Lussac towers so as to recover
as far as possible the nitrous vitriol temporarily taken
from stock.

As the activity of the chemical reactions going on in the
chambers is proportional to the heat produced by them, it is
plain that in a regular normal process the temperature at
the most active and least active zones will bear a constant
ratio to one another, so long as the process is regular ; this
fact affords a very delicate indicator of the regularity of the
process.

§ 28 SULPHURIC ACID 39

36. If a chamber thermometer, placed in the side of the
first chamber about 20 feet from the entrance of the gas
from the Glover tower, that is, in the zone of greatest activ-
ity, registers 100° C, and a thermometer placed in the side
of the second chamber, or a zone of lesser activity, regis-
ters 70° C. , when the process is at its best ami working with
absolute regularity^ the difference between the two readings
represents the relation between the greatest and lesser
activity of that process when normal. If these tempera-
tures vary so as to disturb this difference of 30° C. so little
as 1° C, it is time for the acid maker to investigate his
process and find out what is wrong. This will often enable
him to save serious disturbance in his process before it has
manifested itself in any other way. It must be noted that
it is a disturbance of the difference or ratio, however, and
not of the actual temperatures. The zones of most and
least active reaction ebb and flow slightly in the cham-
bers so that the actual readings of the thermometers may
both be a degree or two higher or lower at various times of
the day and especially at various seasons of the year.

In addition to the temperature readings, the manometer
also affords a delicate test. Manometers registering the
tension of the contents of the first and. last chambers will
show a constant difference of pressure when the process is
regular and constant ; such differejice once determined when
the process is at its best will be maintained so long as nor-
mal conditions prevail.

As a guide to the proper supply of steam at various zones
of the process, drip pans are placed on the sides of the
chambers, which enable a sample of the aqid forming on the
sides of the chambers to be taken and tested with the hydrom-
eter and otherwise examined. This acid, being taken from
the cool sides of the chambers, contains more water than
the average of the acid being formed in the chamber. This
difference is about 3° Baume. A curtain or side drip read-
ing of 50° Baum6 would, therefore, represent approximately
an average formation of 53° Baume acid in that portion of
the chamber.

SULPHURIC ACID

§28

37, Curtain Drip. — For taking these samples the device
shown in Fig. 16 is employed. To the curtain or side lead

mf^

'• !
-M-

ii I
I' !

n I
"I

Ml
Kl
I I
VI
11
|l

I
I
I
I
I
I
I
I
I
I
I

I I

iM^

is attached an inclined lead trough a about 4 feet long. At
its lower end is attached the pipe ^, which passes through the

§ 28 SULPHURIC ACID 41

curtain /, and is bent so as to form a lute or seal. Acid
caught in the trough a runs through the pipe b and drips
into the funnel //, communicating with the hydrometer jar c.
This jar, together with a rack for the hydrometer, etc.,
stands in a lead tray d from whose bottom the drip pipe g
leads to the chamber pan k. As acid is constantly dripping
into h and overflowing from r, the acid in c varies according
to that forming in the chamber, and hence tests of this give
a fair indication of the condition of affairs in the chamber at
that point.

Different acid makers prefer to keep the drips of the dif-
ferent chambers stronger or weaker. This is within cer-
tain limits immaterial. The acid in the first chamber, in
spite of the large amount of water supplied by the Glover
tower, will rarely fall below 52° Baum6. The acid formed
in the chambers should, however, never be allowed to get
strong enough to absorb and retain more nitrous anhydride
than is absolutely inevitable, especially in so far as such
chambers are concerned from which acid is withdrawn from
the system. This strength will also be about 52° Baum6,
and the tendency to absorb nitrogen oxide will increase
with every degree Baum6 above this point. Nor, on the
other hand, must the acid get so weak as to permit
the formation of nitric acid in the chambers. This
strength will be about 45° Baum6. Therefore, the acid
formed in the chambers must in no case be weaker than
44° or 45° Baum6, nor should it be much, if any, stronger
than 52° Baum6.

Although the drips are highly useful adjuncts in con-
trolling the chamber process, samples of the bottom
acid should also be taken at intervals, and in each
individual chamber the acid maker must learn in this
way to compare the actual strength of the acid formed
in the chambers as he finds it with the strength of
the acid as shown by his drip tests. Such tests of the
bottom acid are most satisfactory when taken from a
tank that has been filled with acid drawn off from the
chamber.

42 SULPHURIC ACID § 28

THE PUBIFICATION OF CHAMBER ACID

38. General Remarks. — In addition to the impurities
brought into the process with the burner gas, as was previ-
ously mentioned, some of which will not travel beyond the
Glover tower, chamber acid will contain sulphates of lead
derived from the slow deterioration of the lead apparatus,
and also small quantities of nitrogen oxides and even nitric
acid. From a commercial standpoint, the impurities that are
most injurious are the arsenic and selenium compounds and
even distillation will not entirely eliminate these, unless
special precautions are taken. They will pass over into the
products made from acid contaminated with them (for
example, into muriatic acid and calcined salt cake made
from salt and arsenical sulphuric hydrate). If acid con-
taminated with arsenic or selenium is used for ** pickling"
sheet iron or wire, preparatory to galvanizing or covering
the sheets with zinc, tin, or lead, the galvanic action set up
in the dilute acid bath will cause the arsenic or selenium
to precipitate and become deposited on the iron sheets,
which will prevent the adhesion of the zinc, tin, or lead,
and result in ** blistered '* sheets.

39. Most other impurities, especially lead or iron sul-
phates, will separate in the tanks by sedimentation, or, at
the worst, will produce a discoloration of the acid that does
not unfit the acid for most commercial purposes. Fortu-
nately, very few of the metallic sulphides contain selenium
except in minute traces. Practically all the metallic sul-
phides contain arsenic, and many of these best adapted
otherwise for sulphuric-acid manufacture contain it in con-
siderable quantity. Arsenic, therefore, is the principal
impurity of chamber acid, and on account of its poisonous
characteristics, it becomes especially necessary to eliminate
it. When sulphuric hydrate is used for refining crude
petroleum, or for the manufacture of mixed acid for making
nitroglycerin, arsenic is not detrimental, or at least the
manufacturers do not object to arsenical acids. The arsenic
contained in the enormous quantities of sulphuric hydrate

§ 28 SULPHURIC ACID 43

used in the manufacture of superphosphates and fertilizers
may even be of advantage in destroying insects, etc. ; but
for other purposes, and especially for processes connected
with the manufacture of food products, its elimination
becomes absolutely necessary. If the manufacturer is not
prepared to thoroughly purify his product from arsenic and
intends it for the general market, or for galvanizing,
food products, or other similar purposes, then he must limit
his choice of raw material, often to his great disadvantage
as to cost, to such raw materials as are practically free
from arsenic (as brimstone, some few of the iron bisul-
phides, etc.). If his ores contain only a little arsenic, he
can sometimes obtain a fairly pure acid from the second
chamber, using the acid produced in the Glover tower and
first chamber for purposes less exacting of purity; this,
however, is a dangerous makeshift.

•

40, Purillcation From Arsenic, — As all methods for
the purification of acid from arsenic are based on its precipi-
tation and ultimate removal by sedimentation, it is evident
that this operation must take place when the acid is of least
density ; in other words, while it is still chamber acid (50° to
52° Baum6) and before further concentration.

This statement must, howexer, be qualified in regard to
such manufacturers of 66° Baume and extra-concentrated
acid who are equipped to manufacture such acids by distil-
lation, as will be hereafter described.

In many metallurgical plants, where the acid is a by-
product and the principal value is in the metallic contents
of the metallic sulphides, and in cases where the cheapness
or other advantages outweigh the disadvantage of consider-
able arsenical contents in the raw material, the whole output
must be treated for the elimination of the arsenic.

41, Freiberg: Process for Removing: Arsenic. — Where
this is necessary, the only practical process is a modification
of what is known as the Freiberg: process. This process
depends on the conversion of arsenious oxide into arsenious

44 SULPHURIC ACID § 28

sulphide by means of sulphureted hydrogen gas, the precip-
itation taking place according to the following equation:
As^O^ + 3//,S = As^S^ + ZHfi, As sulphureted hydrogen will
decompose strong sulphuric acid as follows, 3//,5 + H^SO^
= 4:H^0 -\- 25„ it is better to purify the acid as little over
50^ Baum6 as possible. By this process it is stated that at
Freiberg, acid containing as high as .14 per cent, of arsenic
can be purified until it contains only .0002 per cent, of
arsenious oxide As^O^

4:2. In chemical works, where sulphate of ammonia is
prepared from the g-as liquor of illuminating gas works, the
sulphureted hydrogen is a troublesome by-product, but can
be made readily available for purifying the acid in the Frei-
berg process. It contains, however, some pyridine bases
that must first be eliminated if acid of good color is required.
If this source of sulphureted hydrogen is not available, then
it must be prepared by treating iron sulphide with dilute
sulphuric hydrate FeS + H^SO^ = FeSO^ + HJS, The iron
sulphide may be prepared in a simple little furnace by heat-
ing scrap iron or rails with brimstone. On the large scale,
however, it can be very cheaply produced in a cupola fur-
nace by smelting pyrites fines or inferior pyrites with sili-
cious slag.

The iron sulphide so produced is broken into rather large
pieces and filled into a generator, where it is treated with
any available dilute sulphuric acid, such as is often produced
about an acid works from the washings of tanks, tank
cars, etc., and too dirty for general commercial purposes.
These generators are all made on the same general plan
(practically that of Kipp's apparatus, but they are con-
structed out of lead, wood, and iron, and are often made
large enough to hold a charge of iron sulphide sufficient
to last several weeks.

43. Freiberg: Sulphureted Hydrogren Generator. — A

simple and efficient generator for sulphureted hydrogen is
shown in Fig. 17. It consists of a cast-iron generator A with
flanged top and manhole h and an acid reservoir c. This

g 38 SULPHURIC ACID 45

generator, as well as reservoir c, is lined with lead. The
generator is partially filled with iron sulphide d through
the manhole b and the tank with weak sulphuric acid. The
acid will then flow from the reservoir c to the generator A,
and on coming in contact with the iron sulphide will form
sulphureted hydrogen. The valve e and pipe are for carry-
ing away the hydrogen sulphide; when the valve e is open,
the hydrogen sulphide passes constantly away; when e is

closed, the pressure in A rises until the acid is driven back
into the tank c, and the evolution of hydrogen sulphide
practically ceases. The weight of the acid in reservoir c
being carried by the pressure in A, upon opening the valve e
the acid again flows into -4 and generation of gas recom-
mences. A cleaning vent is provided at f, from which the iron
sulphate can be removed when the acid is spent — ^i. e. , entirely
converted into iron sulphate — and^ is a screen of perforated
lead.

46 SULPHURIC ACID § 28

44, Precipitation of tlie Arsenic. — The chamber acid
is then run by gravity into a series of gas-tight lead-lined
boxes or tanks. Each box in the series is provided with a
perforated coil of pipe in the bottom connecting on the out-
side with the main supply pipe for sulphureted hydrogen and
a valve controlling the admission of the gas; it is also con-
nected at the top by means of pipes and valves with every
other box in the series, in such a way that the gas may be
made to pass through any one of the boxes first and then
•consecutively through the others; and, also, that any one of
the boxes may be disconnected temporarily from the series.
In this way, in a series of, say, four boxes, when the acid in
box 1 has had sufficient treatment by the gas, it may be cut
out and boxes 2, 3, and 4 remain. When box 2 has been
treated sufficiently, then boxes 1, 3, and 4 remain in operation.
The box so cut out is allowed to settle as long as necessary.
The precipitation of arsenic sulphide has then taken place to
such an extent that the upper stratum of acid, amounting
to three-quarters or even more of the whole contents, may
be decanted or drawn off by a siphon in a pure state, requir-
ing no further treatment. The rest of the acid containing
the precipitated arsenious sulphide must be filtered.

46, Each series of boxes is provided with two simple
gravity filters, which consist of lead-lined boxes filled with
broken quartz or sand of graduated sizes. The impure acid
is run by means of a pipe and valve on to one of these
filter beds, from which it will emerge practically free from
arsenic. When^one filter becomes foul the other filter is put
into commission and the foul one cleansed by the removal of
the arsenious sulphide from its surface.

The exit gas pipe from the last box of any one or more
series of boxes enters the bottom of the tower shown in the
construction. Just sufficient acid is run into this tower to
prevent the escape of any sulphureted hydrogen that has not
been absorbed in the boxes. The apparatus for the precipi-
tation and filtration of the arsenic sulphide, together with
all pipe connections, is illustrated in Fig. 18 (a) and (d).

§ 28 SULPHURIC ACID 47

The main pipe a brings sulphureted hydrogen from the
generator shown in Fig. 17. The branches and valves ^,,
b^y ^„ and b^ communicate with the gas-tight, lead-lined
boxes C„ C„ C„ and C^, and the perforated coils rf„ etc.

The acid pipe line e is for filling the boxes C„ C„ C„
and C^ with chamber acid by gravity, fitted with branches
and valves ^„ ^„ ^„ and e^.

The return gas pipe / collects the hydrogen sulphide
remaining after it has percolated through the acid in the boxes
and conveys it to tower G, It is fitted with branches and
valves/,,/,,/,, and/.

The tower G is packed in various ways, and a stream of
weak arsenical acid runs down through it, meeting the weak
hydrogen sulphide not taken up by the arsenical acid in
boxes C„ C,, C„ and C^, This stream must be regulated to
completely utilize the hydrogen sulphide and prevent its loss
into the atmosphere. The tower is fitted with acid supply
line A, tank A„ and distributor //,.

The filters / and /, are used alternately. A blow case or
acid ^%% J is used for pumping the purified acid to the storage
tanks.

After a box is suflSciently treated with hydrogen sul-
phide the gas valve is closed and the manhole opened. The
box is then allowed to stand for from 12 to 24 hours, when
the arsenic sulphide will be found to have settled to such an
extent that about three-fourths of the contents of the box
may be decanted off by means of a siphon and passed direct
to storage. The remaining quarter is drawn through pipe k
and branches ^,, ^„ k^^ and k^ into whichever one of the filters
happens to be in commission. This filter strains out the
arsenic sulphide, permitting the purified acid to run through
pipe / into the pumping apparatus, whence it also passes
to storage. The tank is then again filled with acid and
another tank cut out for treatment.

46, StaM Method for Removing Arsenic. — For the

purification from arsenic of comparatively small quantities
of acid, Doctor Stahl's method is very satisfactory. The

48 SULPHURIC ACID g 28

acid is diluted to 40° or 42° Baum6 heated to 80° C, and a
solution of barium sulphide o^ 8.3° Baume is run in at the
bottom of the vessel in such a way that no hydrogen sulphide
escapes. The arsenic trisulphide is filtered off on a sand bed
placed on a layer of quartz lumps, and in this way the arsenic
will be reduced to .01 per cent., but as the acid on standing
in the filter again takes up a little arsenic, it is treated with
gaseous hydrogen sulphide and is thus reduced to .005 per
cent, arsenic.

Arsenic may also be precipitated as a sulphide by means
of the sulphides of sodium, calcium, iron, and ammonium,
and by sodium and barium thiosulphates, but for most pur-
poses these substances are objectionable either on the
ground of cost or because they leave objectionable impurities
dissolved in the acid treated.

CONCENTRATION OF DIIiUTE ACID SOIiUTIONS
AND THE PRODUCTION OF SUIiPHURiC

MONOIIYDRATE

47, The acid solutions resulting from the reactions of
the chamber process consist (1) of chamber acid aver-
aging about 60° Baume, rarely over 62° to 54° Baum6, and
often diluted for purpose of purification as low as 40° Baum6;
(2) of acid concentrated to 60° to 62° Baum6 by the heat of
the burner gas in the Glover tower.

The concentration of these two products varies materially
and must be separately considered.

48, Concentration In Xiead Pans. — The first concentra-
tion of the dilute chamber-acid solutions, varying from
40° to 54° Baum6, which come under the first class above, is
always effected in shallow lead pans. Concentration in lead
can only be made to 60° Baume or slightly over, as the lead
pans are rapidly acted on by hot acid of greater strength.
The evaporation is carried on in these pans by means of
(a) waste heat ; (d) direct heat applied either {c) above or
{d) below the pans, derived from coal, coke, natural or pro-
ducer gas, oil or petroleum, tar, or applied as steam.

§ 28 SULPHURIC ACID *d

Practically, except in special cases, steam is not found sat-
isfactory and the benches used are of two varieties, viz.,
those in which the heat is passed over and those in which
the heat is passed under the pans.

Pans used to be placed over the brimstone burners, utili-
zing the heat of combustion. When pyrites began to take
the place of brimstone, the pans were still placed above the
burners. This practice is now almost entirely done away
with, partly because of the large amount of dust involved
by the use of pyrites and partly because of the trouble
caused by leaks from the pans saturating the costly masonry
of the furnaces with acid and of the difficulty of repairs to
the pans when so placed, but principally because the intro-
duction of the Glover tower utilizes the waste heat of the
furnaces to much better advantage. Fig. 19 includes a pan
bench arranged to be fired from below.

The dilute solution flows continuously through the pan
bench in quantity to insure its leaving the bench a uniform
density of about 60° Baum^. This acid must now be further
concentrated, either in glass, porcelain, or platinum. After
the acid reaches a strength of 64.5° Baum6, it may be further
and finally concentrated in iron stills or the final concentra-
tion may be made in glass or platinum. Below this strength
(64° to 65° Baum6) it acts too strongly on the iron. The
concentration in porcelain cannot be carried beyond about
65.5° Baum6.

49. Concentration In Platinum, or Partly In Plati-
num and Partly In Iron. — In Fig. 19 is shown a bench of
platinum pans or stills /*, o^ and q^ also the bench of lead
pans r, y*, and g^ in which the preliminary concentration is
made.

Platinum stills of circular or oblong shape with rounded
corners are made of many different patterns; some are
provided with platinum covers; some have water-cooled
leaden covers or hoods, as in Fig. 19. The principle, how-
ever, is the same in all; they are practically evaporating
kettles for continuous service, provided with an inlet and

§ 28 SULPHURIC ACID 51

exit for the stream of acid and with means for eliminating
and condensing the steam or weak distillate. During the
gentle evaporation of these dilute hydrates in the lead pans,
little but water, in the shape of steam, is driven off; after the
solution reaches a density of 60° Baume, more and more of
the hydrate is driven off with the water; when the solution
reaches a density of 66° Baum6 (93.5-per-cent. //,5{?J, the
distillate will attain a density as high as 60° Baum6 (77.6-per-
cent. H^SOy When the solution in the pans contains in the
neighborhood of from 95- to 98-per-cent. H^SO^y the distil-
late will have a density of 66° Baum6 (93.5-per-cent. H^SO^,
Much of this distillate is too weak for a reconcentration. It
is sometimes run into the drain, but should be used for dilu-
ting the nitrous vitriol on the Glover tower. The apparatus
shown in Fig. 19 {a) and {b) is continuous in its operation.

The fireplaces a, b, and c communicate with the common
flue d. This flue at one end is arched over with ** pigeon-
hole" or open brickwork, permitting the fire gas to pass
into e\ under and from end to end of a lead pan e. The
heated gas returns under lead pan / through flue f*^ and
then passes through flue g' under lead pan g to the stack.

Chamber acid is run into lead pan g^ whence it flows to
pan f and thence to ^, from which it passes by platinum
pipe // to platinum dish /, covered by a lead water-cooled
hoody. The steam and acid vapors escape by pipe k into
water-cooled condenser / and thence into the small condens-
ing tower M, Acid then flows from platinum dish / by
platinum tube ;/ into platinum dish ^, provided with water-
cooled lead hood and exit to condenser. From platinum
dish o the acid passes through platinum pipe/ into platinum
dish ^, also provided with hood and exit to condenser. As
the acid leaving o will have reached a strength of from
(>4.5° to 65° Baume, an iron dish is often substituted for
platinum dish q. The acid then runs through platinum
pipe r into cooler 5, and thence to storage.

60. Concentratjon in Iron. — Different manufacturers
have different views as to the material best suited to this

SULPHURIC ACID

§28

jT '':■■:

final concentration.
Iron, if properly cast
and of suitable com-
position, is but little
acted on by acid of
64.5'' Baume, and it
is, of course, very
much cheaper than
platinum. On the
other hand, for the
manufacture of the ex-
tra concentrated acid,
from 97- to 98 -per-
cent. jF/^SO, or 79- to
80-per-cent. 5(9, iron
is also more suitable.
Hot acid stronger than
94 - per - cent. //,5(9,
acts strongly on plat-
inum, but has very
little action on iron.
In this country final
concentration in iron
may be said to be the
rule and the practice
is rapidly gaining
ground in Europe.

61 • Concentration
in Glass Retorts or
8tills. — This practice
is practically obsolete
in the United States,
but the following de-
scription of the ap-
paratus sometimes
used will be of interest.
In Fig. 20 (a) is shown

§ 38 SULPHURIC ACID 63

a side view and section of the furnaces and retorts, and
Fig, 20 (6) shows an end view of the same. The glass
retorts c, r,, and c, are arranged in steps as shown. The
acid from the pan bench flows by gravity through the
pipe a and funnel d into the highest retort £. The over-
flow from c flows through the pipe / to c, and so on down
the series; the concentrated acid from the last retort c,
flows to the cooler //, from which it can be drawn by
means of the pipe /. The weak distillate is carried through
the "goosenecks"*/, (/,, and rf, to the vapor Hue r. A
separate fire is maintained under each retort in the fire-
boxesy,y,, andy,. At k, k^, and k^ are the ash-pits The
flue / carries the fire gases to the stack. In case of breakage
of retorts, their contents are carried oflf by means of the
conduit w/.

53, Concentration In Porcelain or Glass Beakers or
I>Ishes: Systems of Negrrier, Webb, Xievlnstein, and
Others. — The principles involved in all these systems of

concentration are very simitar, and, generally speaking, are
merely modifications in details of construction. The acid
flows continuously from dish to dish or beaker to beaker.
The firing is done from below and the acid vapor is carried
away by a separate flue. Fig, 21 shows the Negrier

SULPHURIC ACID

$28

§ 28 SULPHURIC ACID 55

apparatus and illustrates this method of concentration.
All these methods, however, are open to the objection
that it is very difficult to prevent the escape of acid fumes
into the air.

The operation of the Negrier apparatus shown in Fig. 21
is as follows: Pan acid from a flows through conduit b into
the first porcelain dish c^ and so on by means of the lip
on the dishes from one dish to the other ^,, ^, . . . . ^„ until
the strong, concentrated acid reaches the conduit d, through
which it is taken to a cooler and the storage.

Heat is provided by fireplace e. The products of com-
bustion pass under the porcelain dishes Until they reach the
flue / and are carried to the stack. The distillates and
water vapor pass through the flue g and are carried to a
suitable condensing apparatus or to the stack.

53. Concentration by the Kessler Process, — This
method consists of the direct use of heated air or fire gas for
evaporating the water from dilute sulphuric-acid solutions.
The current of hot gas produced from a coke fire or pro-
ducer is brought into immediate contact with the dilute
acid. In this process, the following conditions must be ful-
filled: The current of hot air or gas must be brought into
contact with a sufficiently large surface of acid to imme-
diately and considerably reduce its temperature. The air
or gas must then be completely saturated with steam and
acid vapor. The apparatus must not only be able to resist
the action of hot acid and acid vapors, but must be so con-
structed that the crusts and deposits formed can either be
readily removed or will not interfere with the efficiency of
the apparatus. Under these conditions^ the acid can be con-
centrated at a temperature far below its boiling point. In
order to produce acid of 95-per-cent. H^SO^, boiling at
284° C, the temperature need not exceed 170° to 180° C. ;
for the most highly concentrated acid boiling at 320°, a tem-
perature of 200° to 230° C. will suffice.

54. The Kessler still is shown in detail in Fig. 22 {a)^ (b)^
and {/), Apart from the coke fireplace a, the apparatus is

66 SULPHURIC ACID § 2S

divided into two parts, respectively, the saturator c and the
recuperator d. The hot air enters the saturator at about
300° C. to 450° C. and leaves it at 150° C. The acid mist or
vapor passing out of the saturator is retained in the recu-
perator, which acts as a dephlegmating or distilling
column.

Fig. 22 (a) is a longitudinal section through the whole of
the apparatus. A large coke fire in the furnace a supplies
the hot air that passes through the flue b to the satu-
rator c.

The saturator is constructed of lava (from the town of
Vol vie in France) with deflecting plates in such a way as to
bring the hot gas into close and immediate contact with a
large surface of acid, thus securing immediate reduction in
temperature and saturation of the gas with the steam and
acid vapors formed. The acid vapors contained in the
gases leaving the saturator are recovered in the recupjer-
ator d.

The recuperator d^ shown enlarged in Fig. 22 (r), is a
dephlegmating column, also constructed of Volvic lava. It
is supplied with weak acid. In the recuperator the gas
leaving the saturator at 150° C. is reduced in temperature
to 85° C, at which temperature all the acid vapor contained
in the gas is condensed, while the steam or water vapor
passes out of the apparatus at e. The concentrated acid
passes from the apparatus at /"into the cooler^.

The solutions can be concentrated to 98-per-cent. H^SO^
and Glover tower acid can be used. The fuel used to con-
centrate 100 parts of 95-per-cent. H^SO ^ from 54° Baum6
or 68. 25-per-cent. H^SO^ is stated to be 8 parts of small
gas coke for the hot-air producer and 3 or 4 parts of coal
for power for the exhauster. No weak acid is made, and
the product is clear and free from nitrogen compounds ; no
cooling water is required ; the apparatus takes up little room
and requires little repair.

66. Concentration and Distillation, Startinsr With
tlie Glover Tower. — It has already been stated that the

§ 28 SULPHURIC ACID 57

heat produced in the desulphurizing furnaces is sufficient, if
properly conserved, to concentrate the whole of the acid
made in any chamber plant to 66° Baum6.

This can be done in the Glover tower if the tower is con-
structed so as to stand the action of the hot, concentrated
acid. There are, however, two drawbacks to this plan.
The first is the impure condition of the concentrated acid,
which thus contains most of the impurities of the burner
gas, rendering it fit commercially for only a few purposes, and
the second drawback is the danger of the Glover tower under
these conditions not performing its denitrating function
properly. The latter objection can be overcome in several
ways. Two towers can be placed one above the other, the
burner gas passing from the lower to the upper tower. The
upper tower denitrates the nitrous vitriol and supplies a
stream of hot acid from 58° to 60° Baum^ to the lower
tower, the function of the lower tower being simply one of
concentration. If two chamber systems are near to each
other, as is often the case in a chemical plant, then the
Glover tower of one system may be employed as a deni-
trator and the Glover tower of the other as a concentrator;
the burner gas from the two towers, the one intensely
nitrous an<J the other not nitrous, being thoroughly mixed
with a fan and passed on and distributed by the fan
to the two-chamber systems. .In this case all the nitrous
vitriol is run down the ore tower and denitrated, the result-
mg denitrated acid of 60° to 62° Baume being concentrated
to 66° Baume in the concentrating tower.

The drawback of impurity, however, still remains, and
except when an unusually pure metallic sulphide is used as
raw material, the acid is only fit for limited use.

56. A modification of this plan, however, has now been
in use at several works for some years, producing a very
pure acid at a very low cost. This consists in denitrating
and concentrating the acid in a suitably constructed Glover
tower until it has a density of 64.5° Baum^, at which point,
it will be remembered, hot acid attacks iron but little.

SULPHURIC ACID

This add, with the full
heat imparted to it by
the Glover tower (170° to
200°C.), is run from the
tower directly into a large
cast-iron still (about
8 feet X 3 feet X 6 inches).
This still has a cast-iron
cover and is so set in the
brickwork of the fire
that the fire gas plays all
! around it. In this still
it is rapidly concentrated
to about 95 -per -cent.
H^SO, or some degree of
strength higher than
93.5-per-cent. //,5t?,(B6°
Baumd). The 95-per-
li cent. fffSO^ acid is then
2 run into a connecting
^ iron still, also completely
surrounded with the fire
gases. In this still it is
further concentrated to a
very impure 98-per-cent.
' //,S0,. As nearly all
the 98-per-cent. H^SO^
,acid made in this country
is made for the manufac-
turers of nitroglycerin,
who do not call for a pure
acid, and as after being
mixed with nitric acid to
make the so-called mixed
acid, in which form it is
sold to manufacturers of
nitroglycerin, it is usu-
ally filtered to remove

§ 28 SULPHURIC ACID 59

solid impurities, the impure condition of this acid is of little
moment. The important fact is that the distillates prodiiced
by these two stills, respectively, are pure distillates of 00°
Baum6 and 66*^ Baum6, both of which are commercial solu-
tions largely used in the arts in this country. Furthermore,
as the acid runs hot from the Glover tower to the first iron
still, means are taken to add very small quantities of ammo-
nium-sulphate solution, .1 to .5 per cent, on the 6(r Baum6
acid produced. This not only destroys any nitrogen com-
pounds remaining in the strong, hot acid, but also converts
the volatile arsenious acid into non-volatile arsenic acid,
which therefore either remains in the stills or the 98-per-cent.
concentrate and does not pass over with the distillate of
66° Baum6 and 60° Baum6 acid.

The apparatus employed in this method of concentration
is shown in Fig. 23 (a) and {d). The Glover tower A,
Fig. 23 (6), is connected by the platinum pipe, or nozzle ^,
and the platinum box and tube c with the first iron still ^/.
In this still the acid is concentrated to a strength higher
than 93.5-per-cent. H^SO^, generally to about 95-per-cent.
H^SO^, The distillate from this still will average about
60** Baum6.

The acid from the first still d flows to the second still /
through the pipe e. In this still the acid is concentrated
to 97.5-per-cent. H^SO^. The distillate passing out at g
averages about 66° Baum6. The concentrated acid finds
an outlet through the pipe // into the cooler /. A longi-
tudinal section of one of the stills is shown in Fig. 23 {a),

67. liungre Freezing: Process for the Production of
Snlpliuric Monohydrate. — The solution employed should
contain at least 97-per-cent. H^SO^, and in order to obtain
a good yield of monohydrate should be stronger. The solu-
tion is first cooled and then charged into the iron cells of
an ordinary ice plant. When the solution in the cells is
properly frozen, the cells are dipped in. warm water to
detach the frozen solution from the sides of the cells.
The frozen mass is then crushed and passed to a cast-iron

60 SULPHURIC ACID § 28

centrifugal separator, in which the crystallized mass of
monohydrate is separated from a solution of about 94-per-
cent. H^SO^. The pure crystal monohydrate is then melted
in a water-jacketed enameled pan and run into carboys or
other packages.

68. By the above methods is produced the strongest
acid which it is possible to produce by the chamber process.
For obtaining the monohydrate or stronger solutions of 5(7„
we have already seen that the old Nordhausen process has
been replaced by the contact process.

69. The diagram, Fig. 24, shows the various methods of
manufacturing and concentrating sulphuric acid, and also
the relations of the several processes of manufacture.

A very useful function of the contact process is as an
adjunct to an existing chamber process, where it can be
used for strengthening the solutions of sulphur trioxide pro-
duced in the lead pans or the Glover tower, thus repla-
cing the concentrating plant or enabling a stronger acid to be
produced than is possible by concentration, and at the same,
time increasing the capacity of the plant.

ALKALIES AND
HYDROCHLORIC ACID

(PART 1)

CHEMICAL METHODS

SODIUM CHIiORIBE

OCCURRENCB OF SALT

1. Sodium chloride, or common salt, as the raw

material from which practically all the compounds of sodium
as well as hydrochloric acid, chlorine, and bleaching powder
are more or less directly made, easily stands foremost in its
importance to the human race among the substances occur-
ring in nature. Fortunately it occurs in large quantities in
the ocean, it issues from the earth in many places as brine
from salt springs, and most important of all, it occurs in
large solid beds in almost all countries.

SALT FROM SEA AVATEB

2, The average amount of solid material in the Atlantic
Ocean is about 34 grams per liter, of which a little more
than three-fourths is salt, while the remainder consists

§29

For notice of copyright, see page immediately following the title page.

2 ALKALIES AND HYDROCHLORIC ACID g 29

of chlorides, bromides, iodides, and sulphates of potas-
sium, magnesium, and calcium. The Pacific Ocean con-
tains about the Same amount of solids of approximately
the same composition, while various inland seas range
from comparatively dilute to saturated solutions. Table I
gives the composition of the more important large bodies
of salt water.

TABIjE I

Solid salts

Hfi

Solid Contents:

NaCl

KCl

CaCl^

MgCl,

NaBr \

MgBr^ \

CaSO^

MgSO,

K.SO

CaCO^ ,

Mscc\ s

Atlantic Ocean.
Per Cent.

96.37

77-03
3.89

7.86
1.30

4.63
5- 29

Pacific Ocean.
Per Cent.

3-50
96.50

73- 96

1319
1. 01

4- 63
•3.18

3.85

Mediterranean
w>ea«

Per Cent.

3-37
9<5.63

77.07
2.48

8.76
•49

2.76
8.34

. 10

Salt is obtained from sea water either by evaporating the
water by means of the heat of the sun or by freezing out
the water ; for it would not pay to use fuel for evaporating
such a dilute solution. For this purpose a low, level shore is
selected and a series of basins are hollowed out and lined
with beaten clay, which keeps the water from soaking away.
The brine is kept circulating from one of these basins to the
next until the sun's heat and the hot wind has concentrated

§ 29 ALKALIES AND HYDROCHLORIC ACID 3

it to the crystallization point, when it is allowed to stand
until about 50 per cent, of the salt has crystallized out. The
remainder of the brine, which contains so much magnesium
salts that they would separate out with the salt; is called
bittern. This is run into another vat for the separation of
the potassium and magnesium salts, or it is run back into
the ocean. Salt is produced by this means in this country
in large quantities along Great Salt Lake, Utah, and at a
few places in California. In Europe the principal pro-
duction is in Southern France and Italy; in Siberia con-
siderable salt is obtained by freezing the water instead of
evaporating it.

ROCK SALT

3« The most important source of salt is the large, solid
deposits that have been left by the partial or complete dry-
ing up of inland seas at some prehistoric period. The same
process is going on today at the Dead Sea, the Great Salt
Lake, and other places. These deposits have in the course
of time become covered with a layer of earth that varies
from a few feet to several hundred feet in depth. When
this layer of earth is not too thick, the- salt can be most
economically obtained by running down shafts and mining.
In Louisiana the salt lies only 14 to 16 feet below the sur-
face and is mostly ifiined ; there are also one or two mines
worked in Kansas and in New York, although in these States
the shafts go down 800 feet or more. The most important
and extensive salt mines in the world are at Stassf urt, Ger-
many, which produce not only large quantities of pure salt,
but also the greater part of the world's supply of potassium
salts. The Louisiana rock salt is very pure, but practically
all the rest produced in this country contains iron and other
impurities and has comparatively little sale. The salt is
largely obtained from these deposits by boring down to them,
running in water to form a strong brine, pumping this out,
and treating it like any other brine.

4 ALKALIES AND HYDROCHLORIC ACID § 29

SALT FROM BBINi:

4. Brines may be divided into two classes: Natural
brines, which flow from springs or wells from a natural
reservoir and may be quite dilute ; and artificial brines, which
are made by running water into a rock-salt deposit. These
may always be made saturated if desired. Weak natural
brines are concentrated in some parts of France and Ger-
many by means of a graduator until they are strong enough
to pay for artificial evaporation. This method is now going
out of use, however, and such brines are either worked by
solar evaporation or are discarded altogether. In the United
States the processes used for evaporating brines are the fol-
lowing, named in the order of the number of plants using the
system : Graifters^ solar evaporation , open pan, vacuum pan,
and kettle.

6. Solar f^vaporatlon. — This method depends on the
direct heat of the sun. The brine as it is pumped from the
wells first goes to a settling tank, where the iron, which -is
usually present in the form of acid ferrous carbonate, is
precipitated by the escape of the carbon dioxide and the
oxidizing action of the air as ferric hydroxide. Other sedi-
mentary material also separates out at the same time. The
brine is then run into shallow wooden vats, usually 18 to
20 feet wide, 100 to 400 feet long, and about 8 inches deep,
where it is allowed to stand until salt crystals begin to
separate out, by which time most of the calcium sulphate has
deposited. Finally the concentrated brine goes to the salt
pans, which are similar to the above but not quite so deep.
Here the salt separates as crystals and the brine is renewed
from time to time until a salt layer of about 3 inches is
obtained. The residue of the brine, which contains most of
the chlorides of calcium and magnesium, is then run to waste
and the salt ** harvested " by scraping it together and putting
it into tubs with perforated bottoms, where it is allowed to
thoroughly drain.

The vats are built on piles and arranged so that the brine,
after being pumped into the settling tank, can run by gravity

§ 29 ALKALIES AND HYDROCHLORIC ACID 6

to the other vats. In countries where very little rain falls,
especially during certain seasons of the year, the vats can
stand uncovered continuously. In the eastern part of the
United States, however, where frequent rains occur, it is
necessary to provide the vats with movable covers, which can
be rolled back during fair weather. The salt obtained by
this process is in large, bulky, cubical crystals that occlude
considerable quantities of mother liquor, and on account of
the deliquescent calcium and magnesium chlorides thus
mixed with the salt, it becomes moist in damp weather.

6. Kettle Evaporation. — In the kettle process the brine
is evaporated in cast-iron kettles about 4 feet in diameter by
2 feet deep and heated either by direct fire or a steam jacket.
When necessary for the removal of the iron, the brine is
mixed with a little milk of lime and allowed to settle; it is
then run into kettles and evaporated. The calcium sulphate,
which separates out first, is remoVed from time to time
until the salt begins to crystallize. The salt is removed
from the kettle at intervals, drained in baskets, and then
dumped into bins to thoroughly dry.

When heated by direct fire, the kettles are arranged in
rows of from sixteen to twenty-five over the flues; and as
those at the front end are the hottest, the brine evaporates
most rapidly at that point giving the finest crystals, while
the kettles at the back end produce crystals more like the
solar salt. With steam-jacketed kettles, the product is much
more uniform.

7. The Pan Process. — This is probably the oldest of all
methods using artificial heat, for the Romans at the time of
their occupation of England used practically the same
arrangement as the present, except that their pans were
of lead and only about 6 feet square. The pans a (Fig. 1)
now used are made of iron and are from 70 to 150 feet long,
20 to 25 feet wide, and 12 to 18 inches deep. They are
heated by direct fire. The grates, of which there are three
or four for each pan, with the doors b for charging, are
situated at the front end of the pan and are connected to a

6 ALKALIES AND HYDROCHLORIC ACID § 29

chimney, which is placed at the rear end of each pan by flues
that lead under the pan. The brine, after having been
purified by milk of lime and settled, is led into the back part
of the pan, where it becomes slowly heated and concentrated
so that it deposits its calcium sulphate as it slowly flows
towards the front and hotter portion of the pan, where the

Pio. 1

greater part of the salt is deposited. At intervals the salt
is scraped together and on to the draining boards c by means
of long-handled wooden hoes. The workmen pass between
the pans on the wooden walks d. The roof is cut out at the
peak to allow the steam to escape, but it is covered with a
cap to keep out the rain.

8. Gralners. — An important modification of the pan
process is the so-called grainer. The pan is made of either
iron or wood and of the same general dimensions as the
above, except somewhat deeper. The evaporation is caused
by steam circulating through pipes that are raised about
6 inches above the bottom of the pan and are kept con-
stantly covered with brine ; in other respects the operation
is practically the same as in the pan process.

9. Vacuum Pan Process. — This process leads to a
very fine grade of salt, and on this account is used in several

§ 39 ALKALIES AND HYDROCHLORIC ACID 7

8 ALKALIES AND HYDROCHLORIC ACID § 29

places. Since salt is about equally soluble in hot and cold
water, it is not possible to concentrate the solution in the
pan and then run the solution outside to crystallize, as is
done in many other cases ; but if the vacuum pan is used
for anything more than bringing the brine to its saturation
point, the salt must be allowed to deposit in the pan. This
can, of course, be accomplished by using a simple pan
covered over and partially exhausted, but it is then neces-
sary to open the pan from time to time to remove the salt,
which is an obvious disadvantage. To do away with this
difficulty several continuous-acting vacuum pans have been
proposed, the best of which is Pick's triple-effect evaporator,
shown in Fig. 2.

In this apparatus, the principle is followed of keeping
each element under a less pressure than in the preceding
one, and evaporating its contents by means of steam taken
from the preceding element. The brine enters at g^ and at
r ^ is a vertical coil of pipes, which in the first element is
supplied with steam through e and is sufficiently long to
condense the steam so that it flows as water from the oppo-
site end s. The heat from the steam coil evaporates the
brine dc^ and the steam passes through the pipe /"into a
similar vertical coil at b \ where it condenses and boils the
brine in a' c\ which stands under a less pressure than that
\n ac\ the steam from a! c\ in turn, evaporates the brine in
a!' c'\ which is under a still lower pressure. The salt as it
separates collects in the funnels r, c\ c'\ and can be brought
into the filter chambers rf, d\ d" when desired by turning
the valve at /, /', i". Each filter chamber has a filter in
the bottom portion from which a pipe h returns to the
upper part of the element, so that the mother liquor may be
returned, if desired. The salt may then be washed by
means of the rose x and the wash water run off by the tap/.
The salt can be withdrawn through an opening in the side
of the filter chamber.

10. In preparing fine table salt, the brine is frequently
mixed with sodium carbonate to precipitate, so far as possible.

§ 29 ALKALIES AND HYDROCHLORIC ACID 9

the calcium as carbonate, and then with a little soap, or some
similar substance, to remove the remainder of the calcium
and magnesium as the insoluble soaps of these elements.

In certain European countries, salt used for food is taxed,
but when used for manufacturing purposes the tax is very
light. To prevent fraud, salt to be used for manufacturing
must be denaturated ; that is, rendered unfit for food
purposes. This is accomplished by mixing with the salt
some one of an almost endless number of substances, among
which may be named sodium sulphate, crude petroleum,
coal dust, iron oxide, alum, carbolic acid, etc.

80DIITM CARBONATE

NATURAL, AND ARTIFICIAL. SODA

!!• Natural Occurrence. — Sodium carbonate occurs
in nature widely distributed. It is seldom found, however,
as the normal carbonate, but as a partial decomposition
product of sodium bicarbonate of the composition Na^CO^y
NaHCO^y^Hfiy commonly known as trojia or urao. It has
long been known in Egypt, where it is called Wadi Atrium ^
or Natrium ; in Hungary it is called Szekso, It is also
found in Russia and other countries. Very large deposits
are found in many parts of the United States, especially
in Wyoming and California. In the former State are
found lakes that contain over 2 pounds of crystallized
sodium carbonate per gallon of water and only a small
amount of sodium chloride. Coal is found only 15 miles
away, so that it is estimated that it is possible to make from
98 to 99 per cent, of pure sodium carbonate at one dollar a
ton. A company has recently been incorporated to under-
take its manufacture.

Probably the largest deposits of natural sodium carbonate
in the world occur in California. Mono Lake in that State
has a surface of 65 square miles, and is estimated to contain

10 ALKALIEvS AND HYDROCHLORIC ACID § 29

75,000,000 tons of sodium carbonate and 18,000,000 tons of
sodium bicarbonate. It is high in the mountains, however,
where fuel is scarce and solar evaporation is out of the
question; besides, the difficulty of removing the finished
product makes the working of this deposit impossible, for
the present at any rate. Owen Lake, however, which has
an area of about 110 square miles, has a sodium-carbonate
content of from 40,000,000 to 50,000,000 tons, and is con-
stantly being'added to at the rate of about 200,000 tons each
year. The soda is here obtained by solar evaporation and
considerable quantities are produced. A third large deposit,
which has recently been discovered in Mexico, is about
2^ miles from Adair Bay on the Gulf of California. This
deposit covers an area of about 60 acres to a depth of from
1 to 3 feet, and is only covered by about 3 inches of sandy
silt. The average sample of the dry soda showed 76 per cent,
of sodium carbonate, 5 per cent, of sodium sulphate,
1 per cent, of sodium chloride, and about 18 per cent, of
soluble matter.

The source of the natural soda is probably feldspar rocks
that are decomposed by the atmospheric conditions. The
sodium carbonate formed is washed by rains into lakes,
and, lacking outlets, their waters become supersaturated.
Probably some is also made by the transformation of sodium
chloride to sodium sulphate by calcium or magnesium sul-
phate; then the sodium sulphate is reduced to sodium
sulphide by certain Algae, and the sulphide is converted into
the carbonate.by the action of carbon dioxide.

1*£. Until nearly the end of the 18th century, practically
all the world's supply of soda came from these natural
deposits and from the ashes of certain plants that grow in
or near the sea. The most of the soda came from this latter
source. For this reason, at that time, potassium carbonate,
which is found in the ashes of land plants, was much the
cheaper and more comgionly used alkali. The plant soda
was made in Spain, where it is called barilla ; in France it
is called varil or blanquette.

§ 29 ALKALIES AND HYDROCHLORIC ACID 11

13, Artificial Soda. — The artificial preparation of
sodium carbonate, frequently called soda ashy dates back
to the latter part of the 18th century and has now become
one of the largest of the chemical industries. While a large
number of processes for the manufacture of soda have been
proposed, the only ones at present in use on a large scale
are Le Blanc's process^ the cryolite soda process, the ammo-
nia-soda or Solvay process, and the electrolytic process.
These processes are named in the historical order in which
they became important, but they will be treated in the order
of their present importance in the production of soda ash in
America.

THE SOLVAY PBCK'ESS

14# Historical. — The fact that when solutions of
sodium chloride and ammonium bicarbonate are mixed, a
part of the sodium separates out as sodium bicarbonate, was
probably known in the early part of the last century. It
was not until 1838, however, that it was recognized as a pos-
sible method for the manufacture of sodium carbonate. In
that year H. G. Dyar and J. Hemming took out an English
patent for making sodium carbonate by means of the reaction

NaCl + HNHfiO^ = NaHCO^ + NHjOl

and then heating the sodium bicarbonate to drive off the
carbon dioxide and water, leaving sodium carbonate. This
patent covered the chemistry of the process practically as
it is worked at the present time, and also many of the
mechanical principles. At that time, however, the cost of
ammonia was too great and they did not succeed in keeping
the loss low enough to make the process profitable. About
1855, Schloe^ing and Rolland patented in England some
improvements on the above process and at a factory in
France actually manufactured about 25 tons of soda
a month for nearly 2 years. They did not succeed in
recovering the ammonia sufficiently well, however, and
abandoned the method. Various other inventors worked

12 ALKALIES AND HYDROCHLORIC ACID § 29

upon the process between 1838 and 1863, and fortunes in
time and money were spent to no avail. In the latter year,
Ernest Solvay, a Belgian, took up the process without
knowing much about the other work that had been done
upon it. He worked upon the process until 1873 before the
mechanical difficulties were overcome and the method
became an assured success. From 1873 until the present
time the process has been constantly growing in importance
and strength, so that now far more than half of the world's
supply of soda is made by this method.

15« Outline of the Process. — The process, in brief,
consists in preparing carbon dioxide from limestone, pass-
ing this gas into an ammonium-hydrate solution to form
ammonium bicarbonate, mixing salt solution with the
ammonium bicarbonate, and getting sodium bicarbonate
and ammonium chloride. The sodium bicarbonate is then
calcined to form soda ash and the carbon dioxide is led back
into the process. The ammonium chloride is decomposed by
milk of lime, the ammonia is set free to be used over again,
and the chlorine goes to form calcium chloride, which is
mostly run to waste. The reactions are then

CaCC\ = CaO + CO^
NHpH'\- CO, = HNH^CO^
NaCl+ HNH^CC\ = HNaCO, + NHfil

UfNaCO, = H^O + CO, + Na,CO,
CaO + H,0= Ca(Of/),
Ca{OH\ -f 'ZNH^Cl = 2N//, + CaCl, + 2//, (9

These reactions do not, however, take place in quite so
many steps, for the sodium - chloride and ammonium-
hydroxide* solutions are first mixed and the carbon dioxide
then run in. The reaction between sodium chloride and
ammonium bicarbonate is a reversible ont,^ so that if we
should start with sodium bicarbonate and add ammonium
chloride to it, we would have a certain amount of sodium
chloride and ammonium bicarbonate formed. The reaction
can, therefore, never be complete. It will be driven far-
ther in the desired direction the more of an excess of salt

§ 29 ALKALIES AND HYDROCHLORIC ACID 13

there is present ; and as salt is cheap, it is customary to allow
an excess of salt over the amount necessary to react with
the ammonium bicarbonate. The latter substance is thus
more completely used. Formerly it was very common to
employ solid salt, but this practice is now quite generally
given up and an excess of saturated brine is used.

RAW MATERIALS

16« liimestone. — The nearer pure calcium carbonate
the limestone is naturally, the better it is, but the impuri-
ties in this case are not so serious an objection as in the
Le Blanc soda process. A too high percentage of silica, iron,
or alumina is objectionable, as it causes the limestone to
clinker if the temperature is sufficiently high to burn the
limestone rapidly. When the lime clinkers, dead burns^ it
is almost impossible to slake it and the lime is worthless. A
high percentage of magnesium carbonate is also undesirable
in a limestone, as it lowers the efficiency of the quicklime,
for the magnesium oxide cannot be used to liberate ammonia
from its salts nor to make caustic soda. The limestone
from different parts of the same quarry differs considerably,
as is seen in the following average analyses of the limestone
used by one of the large United States ammonia-soda
works for three consecutive months:

Constituents

October.
Per Cent.

November.
Per Cent.

December.
Per Cent.

SiO^ (insol. in HCL)
Alfi^ and Fefi^ . . .
CaCO

2.95

.80

94.20

2.36

100.31

5.60

.90

83.26

10.41

100. 17

3.95

•30
88.39

7.75
100.39

MzCO

Total

A hard, compact limestone is the most suitable, as,
although it takes a little longer to burn, it gives a quicklime

14 ALKALIES AND HYDROCHLORIC ACID § 29

that is easier to thoroughly slake, and the slaked lime is
usually of a better quality.

17. Brine. — The salt used in this process is in solution,
and the solution may be made from solid salt at the works.
Usually, however, the soda works are so situated that
natural brine or artificial brine, made by dissolving the
rock salt from its bed, can be used. As pure a brine as pos-
sible is desirable, but the ordinary brine used generally con-
tains more or less calcium and magnesium salts, and some-
times iron compounds are also present. The magnesium
salts are the most injurious, for they are not so rapidly
precipitated by the ammonium carbonate in the purification
process (see Art. 21), and the magnesium carbonate sepa-
rates out later when vat liquor is being cooled and is liable
to then stop the conducting pipes. A sample of the Tully
brine used by the Solvay Process Company, at Syracuse,
New York, in 1892, gave the following analysis:

Constituents

Grams per Liter

Sediment

CaSO

.020
4.306
2.718

.250

7.294

CaCl^

M^CL

d *^ «

Total impurities

This same brine contained at that time 292.88 grams of
sodium chloride per liter. This is a good brine, although a
little low in salt.

18. Ammonia. — Although the ammonia is used over
and over in the preparation of ammonia soda, there is,
nevertheless, always more or less of a loss that must be made
up by an addition from outside. The usual sources of ammo-
nia are the coal-gas works and, at the present time, the
by-product coke ovens. This ammonia, from whatever

§ 29 ALKALIES AND HYDROCHLORIC ACID 15

source, comes to the works in the form of crude sulphate,
or frequently as gas liquor, which is a solution of a mixture
of ammonium salts. So long as it contains little or no free
acid, it is acceptable and is purchased strictly on the basis
of the ammonia that it will yield by distillation with lime.
A good gas liquor should contain at least IG per cent, of
ammonia.

19, Ckml and Coke. — Coal is used entirely for the
boilers, and any grade of coal that is suitable for firing
boilers can be used here. Coke is used mostly in the lime
kiln, and should be good oven coke and as free from sulphur
as possible; for with high sulphur, sulphur dioxide is liable
to get into the gas, and at any rate it will yield a lime high
in sulphates. This, if used for making caustic soda, will
cause the formation of large quantities of sodium sulphate
in the caustic liquor, which necessarily means a loss of soda
as well as the necessity of fishing this salt from the caustic
as it is boiled down.

DETAIL8 OF THE PROCESS

20. Carbon Dioxide and Lime. — Since lime is required
for the recovery of the ammonia and carbon dioxide is
necessary for the preparation of the bicarbonate, they are
both best made at the works from limestone. The gas must
be at least 30 per cent, carbon dioxide and, since the
ash of the fuel does not especially interfere in the use of
the lime, the coke, which is the fuel used for burning the
lime, is charged in layers with the limestone.

The most suitable form of lime kiln for use is shown
in Fig. 3. This kiln consists of a shaft from 24 to 40 feet
high and tapering both ways to two-thirds the distance from
the top. The outer shell of iron is lined with firebrick. In
the larger furnaces two rows of brick are used. The whole
kiln is supported by the iron pillars e, which rest on iron
bases set in brickwork. The top of the kiln is provided with
a cover a that can be raised to charge the kiln and is then

16 ALKALIES AND HYDROCHLORIC ACID §29

lowered and rests in a lute of sand or water in 6. At d are
shown 3-inch holes that are ordinarily kept closed with
plugs, but which may be
opened to serve as peep holes
to observe the state of the
kiln, to admit more air if
necessary, and sometimes to
break down the charge. At
/ the kiln is provided with
grate bars that hold the lime
in place until it should be '
removed, when a car is run
under on the track g and
by moving or turning the
bars, the lime emptied into
the car. A platform is
placed at h for the conve-
nience of the workmen using
the peep holes d, and a plat-
form at c for charging pur-
poses. The limestone and
coke are all elevated to the
platform c and then, by

means of cars running on

FlO. s , . • , ...

the track t, the material is

carried to the different kilns and charged in alternate

layers. A pipe for conducting away the gas is provided

about 4 feet from the topof the kiln, the escaping gas being

quite cool at this point, say not over a00° C. The relative

amounts of the limestone and coke that should be charged

must vary at each place, depending on the composition of

the limestone and the coke. The following considerations

will help us in deciding the matter, however.

The best temperature for burning limestone is about

850° C, but if, owing to impurities in the limestone or to

too high ash in. the coke, the limestone tends to fuse at this

temperature, a lower one must be employed and a longer

time spent in the burning. Damp limestone burns at a

§ 29 ALKALIES AND HYDROCHLORIC ACID 17

lower temperature and better than dry limestone, for the
moisture aids the dissociation of the limestone into carbon
dioxide and calcium oxide.

Theoretically considered, 1 kilogram of pure calcium car-
bonate requires 373.5 calories of heat for its decomposition,
and in burning carbon to carbon dioxide, 1 kilogram of car-
bon yields 8,080 calories of heat. Therefore, 1,000 kilograms
of calcium earbonate should be burned by about 46 kilo-
grams of pure carbon. Considering that the escaping gases
carry away heat with them from 1,000 kilograms of calcium
carbonate, we have 440 kilograms of carbon dioxide. Fur-
thermore, the 46 kilograms of carbon will give 169 kilograms
of carbon dioxide, making in all 609 kilograms of carbon
dioxide. It has a specific heat of .22 calory, and if it
escape at 300° C, it will carry with it 609 X 300 X .22
= 40, 194 calories of heat. Then, to burn the carbon, we use
air, which is four-fifths nitrogen, so that the air necessary
to burn 46 kilograms of carbon contains 490f kilograms of
nitrogen. This has a specific heat of .244, and therefore
will carry with it 490f X 300 X .244=35,917 calories.
Therefore, the escaping gases will take a total of 76,011 cal-
ories of heat, which must be supplied by burning more
carbon. This will require about 9.4 kilograms more car-
bon, which in turn will furnish gas to convey heat, and the
amount can be calculated as above. We will then find that
theoretically about 57 kilograms of carbon will burn
1,000 kilograms of calcium carbonate. There is still, how-
ever, to be added in, the loss of heat through radiation from
the sides of the kiln, from the quicklime, which is not quite
cold when drawn, and also the heat required to evaporate
the moisture in the limestone. Taking all of these factors
into consideration, it has been found to be a pretty safe rule
to allow 120 kilograms of pure carbon for every 1,000 kilo-
grams of calcium carbonate. Then, if our limestone is 90 per
cent, calcium carbonate, it will require 1,000 -^ .90 = 1,111.1
kilograms of limestone to give 1,000 kilograms of calcium
carbonate ; and if the coke is only 95 per cent, carbon, it will
require 126.3 kilograms of coke.

18 ALKALIES AND HYDROCHLORIC ACID § 29

Having thus decided upon his charge, the foreman must
watch the results to know if it is right, and he must also
regulate the air supply. He must not allow the tempera-
ture to get too high, or the lime will fuse, dead burn ; nor
fall too low, or too long a time will be required in the burn-
ing. If an insufficient amount of air is supplied, carbon
monoxide will appear in the gas and the air must be
increased ; on the other hand, too much air will show itself
by oxygen in the gas. Ordinarily the supply of air must be
regulated to burn the coke properly and not have an excess.
If the kiln tends to get too hot and dead burn the lime, it is
necessary to reduce the supply of coke. It is often found
necessary to allow part of the limestone to go unburned in
order not to dead burn the rest of the charge and at the
same time avoid carbon monoxide in the lime-kiln gas.

One of the most frequent mechanical difficulties with
which the lime-kiln man must contend is bridging ; that is,
the charge tends to clog at some point of the lower part of
the kiln, and the loose material underneath works out
through the grate (at/. Fig. 3), leaving an arch in the kiln
that prevents the remainder of the charge from feeding
down. When this is observed, it must be remedied at once
by breaking down the arch by means of iron bars. If the
bridge is very low in the kiln, the bars can be inserted from
below; otherwise, they must be used through the peep
holes d. The carbon dioxide, from the limestone and coke,
mixed with the nitrogen of the air, used to burn the coke, is
removed from the kiln by a pipe about 4 feet below the top.
It contains, as especially undesirable impurities, sulphur
dioxide from the sulphur in the coke and considerable dust.
These are removed, as far as possible, by thoroughly washing
the gas before sending it to the carbonating tower. The
scrubber used for this purpose is shown in Fig. 4. The gas
from the kilns enters the scrubber through ^, which, inside of
the apparatus, is perforated its entire length so that the gas
will be uniformly distributed. The gas rises through the
spray of falling water to the first plate r, where it must bubble
through a column of water, then again through the spray to

%-i9 ALKALIES AND HYDROCHLORIC ACID 19

the second plate c, and so on until it passes out through e,
to the carbonating tower. Meanwhile water is admitted
through d in such quantity that it stands at a suitable
height on each plate. Each plate r has a tubey leading to
the next lower one, so that if the water enters too fast or the
holes in c become stopped, the water can overflow through
this pipe. If necessary, the gas can also ascend by this pipe
to the next section of the washer. Finally the wash water
collects in the bottom of the washer and siphons off through rf.

The time as it comes from the kiln is slaked with just
enough water to cause it to crumble, and is then thrown
into a large vat with revolving paddles. In this vat it is
churned with sufficient water to bring it to a specific gravity
of 1. 16, when it is pumped through a screen to remove the
lumps of unburned limestone, and then to the ammonia dis-
tilling: apparatus (see Art. 37).

ALKALIES AND HYDROCHLORIC ACID § 29

21, Purification of tlie Brine. — The brine must be
freed from the calcium, magnesium, and other impurities as
soon as possible after it enters the works. For this purpose
it is used to wash the gases that escape from the ammonia
saturators (see Art. 22), and from the carbonators (see
Art. 23). These waste gases contain ammonia and carbon

dioxide, so that they form ammonium
carbonate in the brine, and precipi-
tate the iron as hydrate and the
calcium and magnesium as car-
bonates.

For washing the gases, coke towers
similar to those used in condensing
hydrochloric acid are sometimes used.
A more suitable style of washer, and
one in much more common use, is
shown in Fig. 5. In this apparatus,
the brine enters at c and slowly over-
flows through corresponding pipes
until it finally passes out at the bot-
tom. Meanwhile, the gases from the
saturator and the carbonator enter
at a under the cap ^, which causes
the gas to spread out and pass
through ^the brine before going to
the next section. The gas finally
passes out at d.

22, Ammonlaeal Brine. — By

washing the waste gas, the brine
receives enough ammonium carbon-
ate to purify it, and must now be
treated with ammonia. This satu-

FlG. 5

ration of the brine with ammonia takes place in an appa-
ratus similar to that shown in Fig. 5, except that not so
many sections are necessary. For saturating enough brine
to make 50 tons of sodium carbonate a day, a saturator made
up of 2 or 3 sections like the above, of 8 feet diameter, and

§ 29 ALKALIES AND HYDROCHLORIC ACID 21

FlO. 6

having a depth of 15 or
18 inches of liquor in each
section, is sufficient. The
brine must be run through
the saturator at such a
rate that it contains from
65 to 70 grams of ammonia
per liter when it leaves the
tower. The ammonia and
ammonium carbonate have
now thrown out the cal-
cium, magnesium, and
iron, and this precipitate
remains suspended in the
liquid, which is run into
the cooling and settling
tanks. The settling ^ vats
are built with a conical
bottom, so that the im-
purities will collect in the
narrow part and may be
drawn off at intervals by
op>ening a valve in the bot-
tom. If the brine does not
settle, it must be filtered,
but usually this will not be
the case. The brine is
cooled in a vat to as low
a temperature as the avail-
able water will cool it, and
should now be clear and
contain 70 grams of ammo-
nia and 270 grams of sodium
chloride per liter.

23. CarbonatlnpT the
Ammonlaeal Brine.

From the settling tanks the

22 ALKALIES AND HYDROCHLORIC ACID g 29

ammoniacal brine goes to the carbonating towers, Fig. 6.
These are iron towers from 60 to 65 feet high an<f about
6 feet in diameter. They are made up of sections, each
about 3^ feet high and bearing iron plates, one at the bottom
and the other one about half way up. Each plate is sur-
mounted with a dome-shaped diaphragm d that is perfo-
rated with a large number of holes.

Between each pair of plates are a number of pipes ^, Fig. 7,
which conduct water to regulate the temperature in the
tower.

The carbonating usually takes place in

two similar towers. In the first, the

ammonium hydrate is converted into

ammonium carbonate and then the brine

is run to the second tower to be finished.

In. this way less ammonia is lost and the

controlling of the temperature is easier.

The temperature in the second tower, especially, must be

very carefully controlled; for if too cold, a fine, muddy

precipitate of sodium bicarbonate is deposited, which is hard

to filter and work with ; while if the temperature is too high,

the yield of sodium bicarbonate is very much diminished.

A temperature is therefore selected that gives the best mean

course between the two difficulties; this temperature is

between 30° and 40° C.

The ammoniacal brine, by standing in the settling tanks,
becomes thoroughly cooled. The gas enters the carbonator
against a pressure of 1} to 2 atmospheres, and in being pumped
against this pressure becomes heated. To make it more
easy to regulate the temperature in the lower part of the
tower, this gas is cooled to about 28° C. before entering
the carbonators. In this way all the materials entering the
towers are thoroughly cooled and the increase in temperature
in the tower is due entirely to the chemical reactions there
taking place. The brine enters the carbonating tower
through the pipe a b. Fig. 6, which enters the tower about
half way down, although a branch of this pipe is provided,
which enters near the top of the tower and may be used

§ 29 ALKALIES AND HYDROCHLORIC ACID 23

when occasion demands. The advantage of introducing the
brine at about the middle of the carbonator is, that the
ammonia has a chance to meet the carbon dioxide sooner
and is converted into carbonate before the top of the tower
is reached. The ammonium carbonate being less volatile
than the ammonia, less ammonia is lost from the carbonator
by this method of working. The carbon-dioxide gas enters
the tower through ^, which is arranged in a rose at the end
so as to distribute the gas uniformly over the bottom of the
tower. This gas, rising through the ammoniacal brine,
converts the ammonia and ammonium carbonate into ammo-
nium bicarbonate, which, in turn, throws out the sodium
bicarbonate in fine crystals. These, for the most part, pass
to the bottom of the tower, in suspension in the liquid, and
flow away through the pipe c,

A small amount of these crystals constantly adhere to the
plates and finally enough collect to clog the holes so much
that the free passage of the gas is interfered with. For this
reason, every 10 days or 2 weeks, it is necessary to empty
the carbonating tower and clean it by blowing in hot water
and steam to dissolve these crystals. The tower must then
be cooled again before use. A number of towers are usually
employed, so that the process does not stop, fresh towers
being brought into use when it is necessary to clean one.

Since the ammonia is the most expensive substance enter-
ing into the process, the effort is constantly made to use it
as completely as possible, even at a sacrifice of other mate-
rials. For this reason, only about two-thirds or three-
fourths of the salt entering the carbonator is converted
into sodium bicarbonate, the remainder being allowed to
remain unchanged in the escaping liquid; a portion of
the carbon dioxide also escapes unused, although the higher
the percentage of carbon dioxide in the gas used, the better
it is utilized.

A rough test to show that the carbonator is working
properly is to draw a cylinder of the liquor as it runs from
the tower and allow it to stand for ^ hour. It should then
have a precipitate of sodium bicarbonarte equal to from

24 ALKALIES AND HYDROCHLORIC ACID § 29

one-third to one-fourth its total volume. The bicarbonate
should be coarse-grained, and when taken from the filters
and crushed in the hand no water should run out of it.

34, Washing tlie Gases. — The gases escaping from the
ammonia saturators contain considerable ammonia and
therefore cannot be allowed to escape directly into the open
air. The gases from the carbonators consist mainly of nitro-
gen, carbon dioxide, and ammonia, and, of course, must also
be washed. The general method of working with these gases
is the same in each case, so that they can most conveniently be
considered together. It has been found that it is an advan-
tage to keep the saturators as well as the washers under a
slightly diminished pressure. Since the ammonia stills con-
nect directly with the saturators, the effect is to give a
reduced pressure in the stills, which causes the ammonia to
be given off more easily and prevents leaks. Fig. 5 shows
a suitable form of washer for this purpose. In order to
avoid all loss of ammonia, so far as possible, two of these
washers are used. For the first washer, brine is used to
absorb the ammonia and carbon dioxide ; the brine then goes
directly to the saturator. The second washer uses as a
wash liquid dilute sulphuric acid, which removes the last
traces of ammonia.

36. Filtration. — The liquor running away from the car-
bonating tower consists of the sodium bicarbonate in sus-
pension and salt, ammonium chloride, and ammonium
bicarbonate in solution. The sodium bicarbonate is sepa-
rated from the mother liquor by vacuum filters or centrif-
ugal machines. Two forms of vacuum filters are in
use ; the older, the so-called sand filter^ consists of a box
about 10 or 15 feet long, 3 feet wide, and about the same
depth. The bottom is perforated and then covered with a
layer of large pebbles, then smaller ones, and finally a coat-
ing of sand. This is covered with a cloth and a series of
slats laid on to protect the filter when the bicarbonate is
shoveled out. The filter is fastened tightly to a large

§ 29 ALKALIES AND HYDROCHLORIC ACID 25

receptacle, from which the air can be exhausted, thus pro-
ducing suction and more rapid filtration. A vacuum of from
one-half to two-thirds of an atmosphere is maintained.
This receiver also serves to catch the mother liquor. Above
each filter is suspended a water pipe that extends the whole
length of the filter and is sufficiently free that it can swing
the width of the trough. This pipe is perforated with fine
holes and enables the workman to easily wash the precipi-
tate. When one of these filters has been filled and the
precipitate washed, it is necessary to shovel out the material
by hand, which requires a number of men.

For this and other reasons another form of filter has been
introduced into many of the most progressive establishments.
This consists of a cylinder about 4 feet long and 3 feet in
diameter, the circumference of which is finely perforated and
covered with cloth. This cylinder revolves in a large trough
filled with the liquor from the carbonating tower; as a
vacuum is maintained on the inside of the cylinder, the
mother liquor passes to the inside and away, while the
sodium bicarbonate is held to the cloth by the outside pres-
sure of the atmosphere. As the cylinder revolves, the por-
tion with the precipitate comes up out of the liquor and
meets a fine spray of water, which thoroughly washes it. It
then passes on until it meets a scraper, which removes it
from the filter and starts it on its way towards the calciner.

Another form of filter, which is somewhat used for the
crude bicarbonate, but more especially for the purified
bicarbonate, is the centrifugal^ which produces a rapid
and complete separation of the mother liquor from the
crystals, but suffers the inconvenience of the sand filter,
that the crystals must be shoveled out by hand.

The centrifugal filter consists of an inner shell, the sides
of which are made of wire gauze or perforated metal and an
outer casing. The inner portion is free to swing about its
axis ; and when a liquid is brought into it, the centrifugal force
throws the contents to the outside, where the solid part
adheres and the liquid passes through to the outer compart-
ment, where it drains off.

26 ALKALIES AND HYDROCHLORIC ACID § 29

The crude bicarbonate from the filters contains consider-
able water, otherwise it is remarkably pure. Its average
composition is

NaHCO^ 70.0 to ?6.0?6

Na^CO, 3.0to 6.0?6

NaCl 3 to .7^

NH, 61*

H,0 20.0 to 26.0*

36. Calcination. — The next step in the process is the
drying of the bicarbonate and its conversion into soda ash;

at the same time, the small amount of ammonia contained
in the crude bicarbonate is driven off and saved. Of the
large number of arrangements for calcining the bicarbo-
nate, only the two most in use will be described here.

§ 29 ALKALIES AND HYDROCHLORIC ACID 27

The pan form of drying and calcining apparatus is shown
in Fig. 8. It consists of an iron pan a covered tightly by an
iron cover d. Through the top of the cover an iron shaft
runs in a gas-tight box and bears the scrapers c. These are
set at an angle to the bottom of the pan, so that when they
revolve they scrape the bicarbonate and carbonate away and
prevent its burning fast, as well as thoroughly mixing the
charge. The pan is heated from the outside by a tire on
the grate/. The damp bicarbonate is charged in through
the door e, which is then closed and the gases escape
through the pipe rf. When the calcination is complete, the
soda ash is withdrawn through the same door e.

A second form of calciner is shown in Fig. 9. This is
superior to the pan form in that it requires comparatively
little labor to operate it. The moist bicarbonate is charged
into the hopper a and is fed into the conveyer rf by the
wheel c and carried forwards by the worm e. At the same
time a suitable amount of calcined soda ash is fed in by the
worm d to keep the bicarbonate in a condition to move.
The mixture is carried forwards to/ where it falls to ^ and
then passes into the iron cylinder /t, which is heated by the

28 ALKALIES AND HYDROCHLORIC ACID § 29

fire from the grate /. The flames from that grate surround
the cylinder and finally go to the chimney through the fluey.
The cylinder h revolves about its long axis on the roll-
ers k. The chain / scrapes the charge loose from the sides
and mixes it. At ;;/ a scoop arrangement is caused to dip
periodically into the charge and bring a portion of it to the
worm ;/, which conveys it outside to carriers. The liberated
gases and vapors pass out through g^ f^ and o,

A modification of the Thelan pan is sometimes used. It
is covered over and the gases escape through a pipe in the
top cover. The scrapers, instead of revolving, move back
and forth over the bottom. It is found most practical in
this apparatus to only drive off the ammonia and three-
fourths of the carbon dioxide and to finish the calcination in
a reverberatory furnace.

The gases from the calciner are passed through condensers
to condense the water and to recover the ammonia as a
solution of ammonium bicarbonate, which is then run to a
special distilling apparatus. The carbon dioxide, from the
decomposition of the sodium bicarbonate, should, theoret-
ically, be almost 100 per cent, pure, and for this reason it
should be especially good for finishing the carbonating of
the ammoniacal brine, but owing to unavoidable leaks in the
apparatus, it is but little better than the lime-kiln gas and
is usually mixed directly with that gas.

27. Ammonia Recovery. — The mother liquor that
comes from the bicarbonate filters contains the greater part
of the ammonia that was contained in the ammoniacal brine;
15 to 20 per cent, of this total ammonia is present as ammo-
nium bicarbonate and the remainder as ammonium chloride.
This mother liquor is run into storage tanks, where enough
gas liquor is added to make up for the loss of ammonia in
the process. The gas liquor contains free ammonia, ammo-
nium sulphate, sulphide, etc. By the addition of this liquor
the solution going to the still is kept as nearly uniform in
composition as possible. Besides the ammonium salts, this
mother liquor contains the sodium chloride from the brine

§ 29 ALKALIES AND HYDROCHLORIC ACID

that is unacted upon, sodium bicarbonate, and small quan-
tities of other salts.

n^ The old system of managing

Jfc ^^ J^ this liquor was to use a com-

T ' mon still and run in a charge

!- -: 1 of liquor and a charge of lime

and then heat. This has, how-
ever, been given up for a con-
tinuous system in practically
all works of importance. A still
of this latter type is shown in
Fig. 10.

The lower part is built up
of wrought-iron rings and is
divided into compartments by
iron plates having a hole in the
center, which is covered with a
hood-shaped piece of iron. The
upper part is built up of cast-
iron sections, also divided by
plates, which serve to break up
the liquor as it passes down
the tower. The liquor to be
distilled comes from the stor-
age tanks and enters the upper
part of the distiller at d and
passes down over the baffle
plates, meeting the ascending
current of hot gases from the
lower part of the apparatus.
In this upper half, which is
called the heater, the ammo-
nium carbonate is decomposed
and driven off, together with
the free ammonia. All the
gases escape through the exit
pipes a, a. At c, a carefully regulated stream of milk of
lime enters and mingles with the descending solution of

KlO. 10

30 ALKALIES AND HYDROCHLORIC ACID § 29

ammonium chloride and other ammonium salts. Steam is
blown in at d through a rose and carries the ammonia set
free by the lime into the upper part of the apparatus, where
it mingles with the other gases and passes out through a.
At e is the waste-liquor outlet, from which the liquor that
is free from ammonia escapes.

The gases from the distiller consist mainly of ammonia
and carbon dioxide saturated with water vapor at 80** or
85® C. and must be cooled and dried before they go to the
saturators for the ammoniacal brines. This is accomplished
by passing the gases through a long pipe coiled in running
water. The gas then passes into the saturators and the
condensed liquor is returned to the still or sent to a special
still along with the condensed liquor from the calciners.

28. Distiller Ijiqaor. — The composition of the liquor
running from the distiller is somewhat variable, depending
on the quality of lime used in making the milk of lime and
on other conditions. It may be stated in general, however,
that it contains as magnesium hydrate or oxide all the
magnesium that was in the lime, for it is found to be
inadvisable to attempt to use little enough lime to utilize
the magnesium oxide, and in the presence of lime it will
not act. It also contains calcium hydrate, calcium car-
bonate, and, principally, calcium chloride and sodium chlo-
ride. The clear liquor does not vary so much. The clear
liquor taken from a series of distillers at the Solvay Process
Company's works at Syracuse in 1897, and used for making
paper filler, had the following composition:

Constituents

Grams per Liter

Constituents

Grams per Liter

CaCl,

NaCl

75 to 85
50 to 75

CaSO

Ca{OH\.,.,

In this distiller waste the chlorine of the salt is lost, and
in addition it occupies valuable land and pollutes streams.

§ 29 ALKALIES AND HYDROCHLORIC ACID 31

The pollution of the streams is, however, not to be com-
pared with that from tank waste (see Art. 83), and this
waste does not suffer decomposition yielding offensive prod-
ucts, as does the other. The best way to dispose of this
waste appears to be to build tight earth walls around an
area and run in the waste. In this way the water and most
of the substances in the solution leach away ; as the lime
becomes carbonated, the residue does comparatively little
damage. Very many efforts have been made to utilize the
waste, or at least to obtain the chlorine contained in it, but
they have met with little success. Also numberless methods
have been proposed for liberating the ammonia in such a
manner that the chlorine would be left in a little more
accessible form, but these also are of but little value.

A small amount of calcium chloride produced by this proc-
ess is used for circulating in pipes in cold-storage and ice
machines, and it has also been utilized somewhat in the man-
ufacture of artificial stone. An important use for it would
be in the manufacture of paper filler, if there were sufficient
demand for the material ; but as compared with the calcium
chloride produced, the demand for paper filler is insignificant.

!39. Ammonia Liost. — When the ammonia-soda process
was first tried, 20 and more parts of ammonium sulphate per
100 parts of sodium carbonate were lost, so that it is small
wonder that it did not pay. This loss has been considerably
reduced, although down to 1890 it was as high as 4 parts of
the sulphate per 100 parts of carbonate. It has since been
steadily reduced until, in England, in 1897, the loss was
about 2 parts per 100, and now in the best-managed works
in this country it is without doubt reduced to from ^ to
2 parts per 100. This loss of ammonia plays a very impor-
tant part in the process, as will be realized if we consider
that ammonium sulphate costs about ten times as much as
sodium carbonate.

30. Properties of Ammonia Soda. — The sodium car-
bonate made by this process is remarkably pure, having an
approximate composition of

32 ALKALIES AND HYDROCHLORIC ACID § 29

Na^CO^ 98.40j^

NaCL 1.28j^

Na^SO^ 07j^

SiO^ 02^

FeX>^ and Alfi^ Olj^

CaCO^ 12j^

MgCO^ 04j^

Some purchasers, having become used to the less pure
Le Blanc soda, even yet demand that sort of soda ash.
This leads the ammonia-soda manufacturer to add salt, or
sodium sulphate, or both, to his ash and sell it as a lower
grade soda ash. The soda ash made by the ammonia-soda
process is of a considerably lower density than that made by
the Le Blanc process, so that in equal bulk we will only have
about 2 parts by weight of ammonia soda to 3 parts by
weight of the Le Blanc soda. For making soda solutions, the
lighter soda dissolves more readily, and for this purpose is
preferred. On the other hand, the denser soda is much to
be preferred for use in furnaces where the charge must be
fused, for it is less easily carried away by the fire gases. The
denser is also better for packing to ship, as it requires much
less space. The light ammonia soda can be concentrated
into the more dense form by calcining in a Mactear or rever-
beratory furnace.

CRYOI^ITE SODA PROCESS

31. In the southern part of Greenland there occurs a
mineral of the composition Na^AlF^, called cryolite, and
so far as known, it does not occur in any quantity in any
other place. In Greenland, however, it is found in large
quantities; the quarry now being worked is 300 feet long
by 150 feet wide and 120 feet deep, and shafts have
been sunk 120 feet farther without showing any sign
of diminution of the supply of material. It can only be
mined in the summer, however, and the short season tends
to limit the output. This mineral was first considered as a
source of soda by Julius Thomsen, a Dane, in the first half of

§ 29 ALKALIES AND HYDROCHLORIC ACID 33

the last century. He developed a method for working the
material, and in 1854 obtained the exclusive right to mine
the cryolite and work it up into sodium carbonate and other
materials in Denmark. He afterwards sold his right to a
company, and in 1865 the Pennsylvania Salt Manufacturing
Company obtained the right to two-thirds of all the cryolite
mined. At the present time there is one soda works in
Denmark using cryolite, but the greater part of the min-
eral brought down is worked up by the American company
at its works at Natrona, Pennsylvania.

The method of working cryolite at the present time is,
even to the furnace used, practically that proposed by
Thomsen 50 years or more ago. The cryolite is first decom-
posed by calcining it with limestone, when the following
reaction takes place:

3/$« The calcination of the mixture takes place in the
reverberatory furnace, which must be of a special construc-
tion, however, for the mixture must be kept at a red heat,
but the temperature must not get so high that the mass
fuses, for the fused mass is very difficult to lixiviate. The
furnace is built with flues under the hearth, so that the
charge can be heated from the bottom as well as from
the top, and the temperature can, by this means, be care-
fully regulated.

33. According to the above reaction, 100 parts of the
cryolite would require 143 parts of calcium carbonate;
but in practice about 150 parts of pure calcium carbonate
are used for 100 parts of cryolite, as the excess renders the
mixture less liable to fusion and increases its porosity when
calcined. Of course, quicklime can be used in place of the
limestone, and at Natrona this is partly done. The mix at
that place is by weight, 100 parts of cryolite, 20 parts
of limestone, and 80 parts of quicklime. A charge for a
furnace is 950 pounds of this mixture, and during calcina-
tion it loses 75 pounds. A charge of this size requires
about 1 hour to finish.

34 ALKALIES AND HYDROCHLORIC ACID § 29

After calcination, the charge is allowed to cool and then
lixiviated. A solution of sodium aluminate is obtained, and
the insoluble calcium fluoride is left in the tank.

34. Calcium Fluoride. — This is of comparatively little
value, although it is used for making hydrofluoric acid and
fluorides of the other metals. It is sometimes used by
glass manufacturers, but must never exceed 6 to 9 per cent,
of the mix, for otherwise too much silica is volatilized and
the silicon tetrafluoride acts on the furnace to too great an
extent. It is also employed as a flux in certain metallur-
gical operations.

35. Sodium Aluminate. — The sodium aluminate is
carbonated by carefully washed lime-kiln gases; sodium
carbonate is left in solution while the aluminum is precipi-
tated as the hydrate. If the carbonation takes place at the
ordinary temperature, the aluminum hydrate separates in
a gelatinous condition, and it is almost impossible to wash
the soda from it. If, however, a suitable higher tempera-
ture is selected, the precipitate obtained is granular and can
be easily filtered and washed on a filter press. The soda
solution is then evaporated and allowed to crystallize. The
crystals are sold as such, dehydrated and sold as soda ash,
or converted into bicarbonate ; they are especially suited for
this latter purpose on account of their high purity. Some-
times the soda solution is converted into caustic soda.

The aluminum hydrate is calcined and sold as aluminum
oxide for the manufacture of metallic aluminum, or is
treated with sulphuric acid for aluminum sulphate or for
alum.

SAIiT CAKE

36. Sodium Sulpliate. — Sodium sulphate occurs natu-
rally in Egypt, Spain, and other European countries, while
in this country it is found in immense deposits in Wyoming
and in some parts of California. It is so extremely cheap,
however, and these deposits are at present so inaccessible that

§ 29 ALKALIES AND HYDROCHLORIC ACID 35

it does not pay to mine them. The native anhydrous sodium
sulphate is called thenardite; the hydrated, vtirabilite.

Sodium sulphate was first described by Glauber in 1658,
although it was probably known before that time He
prepared it by the action of sulphuric acid on salt and
recommended it as a medicine for internal and external use.
He gave it the name sal tnirabile^ and later it was called
sal mirabile Glauberi, The crystallized salt is even yet
called Glauber* s salt.

The manufacture of this substance, which, when artificially
prepared, is usually known as salt cake, depends almost
entirely on the reaction between sodium chloride and sul-
phuric acid. The latter may be used ready made or formed at
the instant of its action. In the first case, acid sodium sul-
phate is first formed ; afterwards, the normal salt, so that
the reactions are :

NaCl-\- H^SO, = NaHSO^ + HCl
NaCl+NaHSO^ = Na^SO^ -f HCl

In the second method, instead of sulphuric acid, sulphur
dioxide, oxygen, and steam are brought together with the
salt, giving the reaction

maCl + 2SO^ + 2//^0 + 0, = 2JVa^SO, + ^HCl

The first method is the older and at the same time the
most used process for making salt cake.

CRUDE MATBRIAL.S

37. Salt. — The kind of salt best suited to the making
of salt cake is what is known in this country as cattle salt.
The coarse crystals form a spongy mass that readily absorbs
the acid and aids the decomposition in this way. The fine-
grained, so-called, table salt is totally unsuited for this
purpose.

The salt as it comes to the works usually contains
^bout 95 per cent, of sodium chloride and about 5 per cent,
of water, with other minor impurities.

86 ALKALIES AND HYDROCHLORIC ACID § 29

38. Salphiirlc Acid. — The ordinary impurities occur-
ring in sulphuric acid are usually of very little importance
in this connection. When the salt cake is to be used for
glass making, the iron in the acid should be kept as low as
possible, and arsenic, on account of its getting into the
hydrochloric acid, is sometimes objectionable. The concen-
tration of the acid should be 60° or 60.4° Baume. Weaker
acid is not good on account of its acting strongly on the
decomposing pans, causing slow work and weak hydrochloric
acid. Acid as weak as 55.5° Baum6 may be used, although
it is undesirable, and weaker than this should never be
tolerated. An acid stronger than 60.4° Baume, on the other
hand, causes a too rapid evolution of the hydrochloric acid.

APPARATUS AND METHOD OF MAKTTFACTUItB

39. The apparatus used in the manufacture of salt cake
varies considerably in detail, but according to its essential
features may be divided into ofien roasters, blind roasters or
muffles, and mechanical furnaces.

40. Open Roasters.— The open roaster shown in Fig. 11
consists of two parts, the/rtw a and the roaster b.

41, The Pan. — Since the pan must stand the action of
sulphuric acid, it was nt first assumed that it must be made
of lead, but this material has the derided disadvantage of
soon wearing out by the action of the tools used in mixing

§ 29 ALKALIES AND HYDROCHLORIC ACID 37

the salt and acid, and in transferring the product to the
roaster b. A very low and carefully regulated temperature
must also be employed on account of the low melting point
of the lead. Lead has, therefore, been almost entirely dis-
carded in favor of iron for pans, although even now, where
it is desired to make a salt cake very free from iron, lead
pans are used. The iron pans are from 9 to 11 feet in
diameter and from 1 foot 9 inches to 2 feet 6 inches in depth.
They are made about 6 inches thick on the bottom and taper
to about 2 inches at the edge,. and are covered with a brick
arch with an outlet pipe c for the escape of the hydrochloric
acid. The pans are supported by their edges by supporting
walls, and are heated by direct fire from a grate d^ which
is covered by a section of an arch to spread the flame and
prevent overheating in one place and so burning the iron.
Since the pans are heated nearly or quite to redness, when
the batch is transferred to the roaster and, in rapid work, a
new charge of, possibly damp, salt introduced before the pan
has cooled very much, they must be able to withstand consid-
erable temperature changes, as well as the action of the acids.
Between the pan and the roaster is a slide r, which is best
made of two thin sheets of iron placed a few inches apart
with a packing of salt, to keep the hydrochloric acid from the
pan separate from that of the roaster. By this means the
condensation of the pan acid is easier and a much purer acid
is obtained than would be gotten from the mixed gases. The
connection between the pan and roaster is only kept open long
enough to permit of the transfer of the batch of salt cake.

42. Management of the Pan. — Salt to the amount of
from 600 to 1,000 pounds, depending on the preference of
the management, but usually 800 to 900 pounds, is shoveled
into the pan through the working door, and then enough sul-
phuric acid, of 60° to 60.4° Baum6 (taken cold), having been
previously heated, is run in through a pipe in the cover of
the pan and the mixture heated.

The amount of sulphuric acid used is naturally regulated by
the charge of salt and the moisture in the salt. Theoretically,

38 ALKALIES AND HYDROCHLORIC ACID § 29

every 68.5 parts, or pounds, of salt should have 49 parts,
or pounds, of sulphuric acid H^SO^\ that is, every 100 parts,
by weight, of TV^C/ requires 83. 76 parts, by weight, oiH^SO^,
Sulphuric acid of 60° Baum6 is 78-per-cent. H^SO^y and there-
fore 100 parts of NaCl requires 107.37 parts of sulphuric
acid of 60° Baum6. Since, however, the salt used is only
about 96 per cent. NaCl^ the amount of 60° Baum6 acid will
be 102 parts, by weight, for every 100 parts, by weight, of
salt. Some allowance must, however, be made for loss of
sulphuric acid, by volatilization, in the pan and roaster, so
that in most works, for making strong salt cake, about
2^^ parts, by weight, of sulphuric acid in excess of the amount
calculated is added for each 100 parts, by weight, of salt.
The practice, then, is to add 104.5 parts, by weight, of
60° Baum6 sulphuric acid to each 100 parts, by weight, of
salt charged. If weaker acid is used, the calculation of the
amount of acid can be carried out in the same way.

The charge of acid is never weighed, but is measured so
that it must be added each time at the same temperature.
The salt and acid are analyzed daily in the laboratory and
tables are furnished the pan man, so that by determining
the specific gravity of the acid coming to him, at a constant
temperature, he can easily determine the amount of acid to
add. The be$t temperature for the acid is a matter of
opinion, but it should never be below 50° C, while some use
it at nearly 100° C. The hotter the acid, the less it acts on
the pan ; but with too hot acid, the hydrochloric acid is given
off too rapidly and it is difficult to condense it, while a
thorough mixing of the batch is almost impossible. An acid
of about 60° C. is considered the best.

Under the best conditions of working, the batch in the pan
foams badly and has a tendency to foam over. This diffi-
culty can be quite largely met by adding a small piece of
paraffin as soon as the sulphuric acid is run in.

As soon as the acid is added to the salt the mixture is
thoroughly stirred by the pan man, for which purpose he
uses a long-handled iron rake inserted through a hole in the
working door. At best, considerable hydrochloric-acid gas

§ 29 ALKALIES AND HYDROCHLORIC ACID 39

escapes during this operation, but by heaping salt about
the handle of the rake where it passes through the door,
the escape of the gas is reduced as much as possible. When
the mixture has been brought to the consistency of thin mud
and all the lumps of salt have been broken, the rake is with-
drawn and the door closed as tightly as possible by piling salt
against it. The door itself is made of slate or of lead-cov-
ered cast iron and is set in a frame of acid-resisting stone.
The workroom should be thoroughly ventilated to relieve the
workmen, so far as possible, from the inconvenience of the
acid that unavoidably escapes.

In this operation in the pan, the first half of the reaction
takes place and, theoretically, 50 per cent, of the total
hydrochloric acid is evolved. Practically, the heating of the
pan is continued until about 70 per cent, of the total hydro-
chloric acid is given off, for it is advisable to have as much of
the hydrochloric acid evolved in the pan as possible. The pan
.hydrochloric acid is purer and easier to condense than that
from the roaster. The batch in the pan is considered
finished when the mixture offers considerable resistance to
the moving backwards and forwards of the rake, owing to
the stiffness of the mass. The finishing of the batch then
requires a higher heat than can be obtained in the pan.
Assuming the roaster bed to be empty, at a bright red heat,
and the batch in the pan finished, the slide e is raised, the
pan door opened, and the pan man, by means of a long-
handled, spoon-like shovel, transfers the charge to the roaster,
where it is at once spread out evenly by the roaster man.
There is always a tendency for the acid salt cake to stick to the
pan, especially if it is not set so as to be evenly heated. This
is best remedied by care in setting the pans so that the heat-
ing will be uniform ; where such cakes do form, they should
be removed before adding a new charge, otherwise the pan
is very likely to crack.

43. Open Roaster. — The open roaster b, shown in Fig. 11,
consists of a shallow basin from 12 to 15 feet long and nar-
row enough for the batch to be handled by the workmen

40 ALKALIES AND HYDROCHLORIC ACID § 29

using long-handled hoes. It is simply a form of reverbera-
tory furnace and is lined with carefully placed firebricks.
The material is heated by direct flame from a coke fire on
the grate f and the products of combustion, together
with the hydrochloric acid, escape through the pipe g.
Since all the fire gases mix with the hydrochloric acid
in the open roaster, it is very much diluted and its
complete condensation to a strong acid solution is very
difficult.

44, Managrement of the Roaster. — The batch is spread
evenly over the bed of the roaster and at intervals of from
10 to 15 minutes must be turned over and all lumps broken.
For this purpose the furnace man uses a wrought-iron rake
and a bar of wrought iron flattened at the end into a blade.
The tools are introduced into the furnace through the doors h
and /, and are suspended from hooks hanging from the ceil-
ing in front of the furnace door. By thus suspending the
tools, part of their weight is taken off from the furnace man,
but the work is hard and disagreeable at best. The furnace
must be kept hot to get the batch off as quickly as possible,
but it must not be allowed to get too hot or the batch will
flux. A small amount of fluxing can be taken care of and
the lumps broken, but if it once gets ahead of the furnace man,
especially next the fire-bridge k^ it is almost fatal to the
charge, for it cannot be controlled and the salt cake is then
almost useless for the black-ash furnace. The way to avoid
this fluxing is to carefully watch the fire.

The furnace work is finished when no more vapors are
given off, even on turning the batch and when it is quite red
hot, but it must never flux at any point. The salt cake is
then drawn by means of wrought-iron hoes into steel bar-
rows and carried to the storeroom. The hydrochloric acid
given off in the open roaster is mixed with the gases and dust
from the grate, so that its condensation to a strong acid is
difficult and there is danger of the condensers becoming
stopped by the dust. To obviate this difficulty, the blind^ or
muffle, roaster has been adopted by many manufacturers.

§ 29 ALKALIES AND HYDROCHLORIC ACID 41

45. Blind, or Muffle, Roaster. — This roaster employs a
pan of practically the same dimensions and setting as the open
roaster, and it is sometimes heated by the waste gases from
the muffle heating. It is better, however, to heat it by its
own fire, as in the preceding case, for although it saves some
fuel when waste heat is employed, direct firing makes the
working of the pan independent of the muffle, which is in
many cases a decided advantage. The essential difference
between this method and the preceding one is in the roaster.
Here, instead of having the batch heated by the direct fire,
with its numerous disadvantages, the batch is brought into
a closed muffle and heated by the heat conducted by the
muffle walls from the outside flues. The muffle walls are
made of brick and must be quite thin, or it will not be pos-
sible to get the charge sufficiently hot. Since the walls are
thin, they are liable to be damaged by the tools used in
working the material, or they may crack on account of the
temperature changes. Since the pressure inside the muffle
is greater than on the outside, if such a crack forms, large
quantities of hydrochloric acid may escape into the chimney
gases and great damage be done before the leak is discovered.
These difficulties led Deacon to devise his plus-pressure
furnace.

46. Deacon's Plus-Pressure Furnace. — In the muf-
fle roaster just described, the fire-grate is nearly on the same
level as the muffle, and a draft is produced by means of a
chimney; so that necessarily the flues about the muffle are
under diminished pressure, while, on account of the acid-
absorption apparatus, the acid in the muffle is under greater
pressure than the atmosphere. The result of this is,
therefore, that if there is a leak in the muffle, the hydro-
chloric acid will escape into the chimney. Deacon reverses
this condition, not by diminishing the pressure in the muffle,
but by increasing the pressure in the flues by putting the
fire-grate ^, Fig. 12, much lower than the muffle c. The
hot gases rising in the vertical flue to the muffle flues b, b,
produce a pressure on the latter, so that if there is a leak in

42 ALKALIES AND HYDROCHLORIC ACID % 29

the muffle, the fliie gases go in and do comparatively little
harm. It is practically putting the muffle at the top of the
chimney instead of at the bottom, as in the other style. As
shown in the illustration, the fire gases rise from the grate a,
pass over the muffle and then through a series of flues on
the under side of the muffle, and finally go to heat the pan
or go direct to the chimney, as the manufacturer prefers.

47. All that has been said about the working of open
roasters applies equally well to the muflle furnaces. The
heating of the whole furnace bed is more uniform and the
danger of overheating is not so great.

The advantages of the two styles of roaster may be sum-
marized as follows: The open roaster works more rapidly
because the charge can be got hotter, and therefore gives a
large yield of salt cake. For the same reason, it is possi-
ble to make a stronger salt cake, i. e., one containing a
higher percentage of normal sodium sulphate. Less repairs
are needed, and it is impossible for the acid to accidentally
escape anywhere except through the condensers. The
muffle roaster, on the other hand, makes possible a better
condensation of the hydrochloric acid, and therefore pro-
duces a cheaper and stronger acid. It requires less acid
per unit of salt, and coal, instead of the more expensive

g 29 ALKALIES AND HYDROCHLORIC ACID 43

coke, can be used for firing. More fuel is required,
however, so that the last item probably does not represent

much saving. The advantages of the two systems are
so evenly divided that some firms prefer one system and

44 ALKALIES AND HYDROCHLORIC ACID § 29

some the other, so that the two systems are about equally

used.

■

48, Mechanical Furnaces. — In the preceding methods
of working, the batch must be transferred from the pan to
the roaster and carefully worked to prevent the leaving, in
the finished product, of lumps -of salt unacted upon. It
requires a certain amount of skill to do this properly,
and so the manufacturer is to a certain extent in the hands
of his workmen ; furthermore, every time the furnace door
is opened, acid gas escapes into the room and produces an
unhealthy atmosphere for the workmen. These consid-
erations have led to various attempts to perform all this
work mechanically, but the only arrangement that is
commercially successful is the Mactear furnace, shown in
Fig. 13.

This furnace consists of the pan a in the center of the
ijiovable hearth b, and is heated by the gas from the grate c.
The salt is fed in continuously through the hopper d^
and at the same time the proper amount of acid flows in
through e. The two substances mix and partly react in the
pan, and then the mixture is slowly worked over on to the
hearth by the stirrer /. The hearth revolves on small
wheels running on the tracks g, and by this motion and the
stirrers extending from / to the outer edge and turned by
the outside cogs, as shown, the charge is worked to the
outer edge by the time the reaction is completed. The salt
cake then flows into the annular trough h^ by means of
which it is conducted from the apparatus. All the joints
of the apparatus are closed by aprons dipping into lutes of
molten sulphate, but even this does not altogether protect
the outside from the acid fumes.

These furnaces have the advantages that they do away
with a large amount of manual labor, yield a continuous
product, and allow the hydrochloric acid to be more easily
condensed, for it comes in a continuous, uniform stream,
while in the hand furnace the evolution of acid is variable.
But they have the disadvantage that the hydrochloric

§ 29 ALKALIES AND HYDROCHLORIC ACID 45

acid cannot be made so strong as with hand work, and
the machinery is expensive and requires a large amount
of repairs. These disadvantages have restricted the use
of this furnace, so that probably not over 15 per cent, of
the salt cake made at the present time is made by
them.

49. Yield of Salt Cake.— The yield of salt cake will
naturally differ in different works and with different appa-
ratus, but the amount that may be expected with good
work, etc. will be about as follows: 100 parts, by weight, of
pure salt should, theoretically, yield 121.5 parts, by weight,
of salt cake. As already pointed out, the salt used rarely
contains over 95 per cent, of sodium chloride; and, of
course, this must lower the yield of salt cake. If, however,
we adopt the rational method of calculating the percentage
yield on the sodium chloride actually used, the yield should
be very nearly theoretical. Works are in operation that
produce 121.2 parts of salt cake for 100 parts of pure chlo-
ride used.

50. Properties of Salt Cake. — A good quality of salt
cake should be finely granular and yellowish white, or bet-
ter, pure white in color. A deep yellow or reddish-brown
color shows much iron, while a dirty-gray color indicates
incomplete decomposition of the salt. The salt cake should
not contain over 1 per cent, of free sulphuric acid, nor more
than .6 per cent, of sodium chloride. When intended for
use in glass manufacture, the iron should not exceed .2 per
cent. Fefi^,

51. Uses for Salt Cake. — Sodium sulphate is most
largely used in making sodium carbonate by the LeBlanc
process. It is also used in making glass and ultramarine,
and in dyeing and coloring. It finds a smaller use in making
sodium acetate and other sodium salts from the correspond-
ing calcium salts, and the crystallized sodium sulphate
(Glauber's salts) is used in medicine.

46 ALKALIES AND HYDROCHLORIC ACID § 29

SODA BY THE liB BliANC PROCESS

52. lie Blanc^s process for making sodium carbonate
from salt consists in first making sodium sulphate, as
already described, and then converting this into sodium car-
bonate by fusing the sulphate with a mixture of calcium
carbonate and carbon. De la M6therie had previously pro-
posed heating sodium sulphate with carbon to reduce it and
convert it into the carbonate, so the point especially made
by Le Blanc was the introduction of calcium carbonate into
the mixture, and this was the important step that made the
process a commercial success.

The process comprehends the starting with sulphur in its
elementary form, or the much less valuable iron pyrites,
with calcium carbonate, carbon, and sodium chloride as raw
materials, and ending with the sodium as carbonate, the
chlorine free or as hydrochloric acid, the calcium carbonate
as when starting, and the sulphur free, so that the only
material used up is the carbon, and there are no by-products.
It is not possible to realize this condition entirely, however,
and it is only comparatively recently that it has been possi-
ble to recover the sulphur commercially.

This round can be represented by the following reactions:

25 + 3(9, + %Hfi = 2//,S0^
^Naa+ 2H^S0^ = %Na^SO^ + 4.HCI

2Na^SO, + 4:C= %Na^S + 4C(9.
^Na^S + %CaCO, = %Na^CO, + %CaS
2CaS + 2C0, + %Hfi = ^CaCO, + %H^S
%H^S + 4(9, = %H^SO,
or 2iy.5 + ^, = 25 + 2Hfi

Or, combining them, we get

%NaCl + C + (9, + Hfi = Na^CO^ + %HCl

That is, theoretically, for 117 parts, by weight, of salt,
only 12 parts, by weight, of carbon are required to convert
it into sodium carbonate and hydrochloric acid, which makes
it apparently a cheap and simple process. The practice is,
however, not nearly so fine, for actually 400 to 500 parts of

§ 29 ALKALIES AND HYDROCHLORIC ACID 47

carbon are required to every 117 parts of salt. In addition
to this, there is a large amount of money invested in the
plant and constantly required for labor and repairs ; besides,
the reactions do not go as smoothly as represented.

53. The reaction that takes place when carbon, sodium
sulphate, and calcium carbonate are fused together has been
the subject of almost endless discussion, especially with
regard to the calcium compound, for it is well known that
an insoluble calcium sulphide is not formed with either
hydrogen sulphide or ammonium sulphide; therefore, it was
long held that the calcium compound formed in the above
must be an oxysulphide CaO^CaS, It is impossible to go
into a discussion of this subject, but it may be taken as
definitely settled that the reactions take place practically as
represented above, the calcium sulphide formed being insol-
uble. At the end of the operation the reaction

CaCO^ + C = CaO + 2C(?

begins, and serves as a signal for the withdrawing of the
charge, for the carbon monoxide comes up through the
material and burns with long, pointed flames, called candles^
and thus indicates that the transformation is complete.
This reaction continues for a long time after the charge is
withdrawn and while it is cooling, so that the escaping gas
leaves the material porous, and for that reason much easier
to lixiviate in a later stage of the work.

RAW MATERTAL.S

54. Sodium Snlpliate. — As the preparation of salt cake
has already been described, we will avoid repetition by con-
sidering it here as one of the raw materials. The sodium
sulphate should be fine and porous, not fluxed, and should
contain 96 or 97 per cent, of sodium sulphate. It is better if
it contains a little free acid, as this lessens the probability
of its containing much salt. The acid should not, however,
exceed 2 per cent. , and the salt not over ^ or 1 per cent.

48 ALKALIES AND HYDROCHLORIC ACID § 29

55. Calcium Carbonate. — The calcium carbonate is
usually chalk or high-grade limestone. All impurities are
bad, and magnesium and silica are especially so because
they form insoluble compounds containing sodium and so
cause a loss of sodium compounds. The limestone is crushed
to the size of a pea or bean before being used, but does not
need to be fine, and is better if not too fine. Caustic mud
(see Art. 97) and calcium carbonate from the sulphur
recovery (see Art. 82, et seg,) are also sometimes used, but
they are so light that they dp not flux well.

56. Carbon. — The carbon is supplied in the form of
powdered coat, which should be low in ash, not over 7 per
cent, being allowable, and one that gives a high yield of
coke. The presence of a moderate amount of pyrites does
not interfere, but the less nitrogen present the better, for it
leads to the formation of cyanides, cyanates, and ferrocya-
nides, the latter introducing iron into the ash.

DETAILS OF THE PROCESS

57. The mixture varies considerably in the proportions
of the constituents, probably partly on account of impurities
in the coal and limestone, but even taking that into account
there is a wide variation, each works using the mix that it
considers gives the best result. The theoretically correct
proportions can, of course, be calculated from the reactions
given in Art. 52. Leaving out the reaction

CaCO^ +C= CaO + 2C6>

the proportion will be 100 pounds of salt cake, 70 pounds of
calcium carbonate, and 17 pounds of carbon; taking thi?
reaction into account, it will be approximately 100 pounds of
salt cake, 75 pounds of limestone, and 20 pounds of carbon.
In practice much more coal is required, for some of it burns
and some is left in the product. On account of this coal
that remains in the flux, the fused mixture is black and is
called black ash.

§29 ALKALIES AND HYDROCHLORIC ACID 49

In the hand-worked furnaces about an average mixture is
100 pounds of salt cake, 9t:i pounds of good limestone, and
48 pounds of coal, but in the mechanical furnaces, which are
now largely used, the charge is frequently cut down to as
low as 100 pounds of salt cake, 80 pounds of limestone, and
30 pounds of coal.

58. Hand Furnaces. — These are simply reverberatory
furnaces adapted to this special purpose. Fig, 14 shows a
front elevation and vertical and horizontal sections of one of
these furnaces. The fire-grate is at a and the hot gases
pass over the bridge g on to the bed of the furnace b c,
which is divided into two sections, and then over the liquid

50 ALKALIES AND HYDROCHLORIC ACID § 29

to be evaporated in the pan d. The fire-bridge g is built
with a flue //, which permits the air to circulate freely, thus
keeping the bridge cool and retarding its burning out. The
bed of the furnace is usually about 15 feet long by 7 feet
wide, and a charge of about 700 pounds, more or less, of the
mixture is worked at a time.

59. Management of the Furnace. — The charge is first
introduced on to the back half of the furnace, through the
hopper ^, and is spread out and allowed to get hot and dry,
being occasionally turned. When it is thoroughly heated
and the front part of the furnace is hot, the charge is trans-
ferred to this part of the furnace and a new charge intro-
duced in the back. The principal part of the making of the
black ash takes place on this front bed of the furnace, and
here also the work and skill of the furnace man comes into
play. Very soon after the mixture is brought on to the
working bed of the furnace it begins to melt in places; then
the furnace man must turn the mixture so that the melted
portion of the material is turned under and the under part
comes to the top. By working the mixture in this way, the
furnace man must gradually thoroughly mix the whole mass
of material and bring it to a rather soft state of fusion.
This requires an almost white heat, and to get up the tem-
perature as well as to rest himself, the furnace man up to
this point only works for a few minutes at a time and then
closes the furnace door for about 10 minutes before mixing
again. The chemical action only begins when the mixture
is in a state of pasty fusion (it never gets past the pasty
stage), and when this condition is reached the reaction must
be finished as quickly as possible. The furnace man is busy
from now on, stirring and mixing the mass and working it
towards the door of the furnace. When the reaction is com-
pleted, flames of carbon monoxide, colored yellow by the
sodium (so-called candles), will appear and the black ash is
then worked out into a barrow. The proper time must be
selected for ** balling " together and withdrawing the charge,
for otherwise it will be underdone or overdone. If not

§ 29 ALKALIES AND HYDROCHLORIC ACID 51

allowed to remain in the furnace long enough, it will contain
unchanged sodium sulphate, and also be dense and hard to
lixiviate ; when in this condition it is called soft ball, for the
last reaction, which gives the gas and causes the porosity,
has not had an opportunity to start. On the other hand, if
left too long, the gas of this last reaction will escape while
the material is still in a soft condition and it will then settle
into a hard mass, burned ball, which is difficult to lixiviate.
Under proper working, however, the material is balled
together when candles appear, and it is brought into an iron
barrow, where, by the continued action between the carbon
and the limestone, gas continuously escapes as the material
cools, and so leaves it porous; the slaking of the lime so
formed assists in the lixiviation. The principal difficulties
occurring In the black-ash furnace are the forming of these
**soft" or ** burned" balls, and the avoiding of them
depends almost entirely on the furnace man. The way to
avoid them is to have the furnace hot, keep the batch well
mixed, and to bring the temperature well up at the end of
the work; then, with proper judgment as to the time to
withdraw the charge, good results are not difficult to obtain.

60. Mechanical Furnaces. — Although the tools are
suspended by chains and hooks, the continuous handling of
them at the high temperature that exists is very hard for
the workmen, and much depends on the good will of the work-
man to get a good result. For these reasons, and to save the
cost of the expensive hand labor, mechanical furnaces are
very desirable. The first furnaces of this kind that were
tried were very expensive to operate on account of the
frequent repairs made necessary by the great wear and tear.
Furthermore, it was difficult to watch for the candles and
draw at the proper time to avoid overburned ash. The
excessive repairs were finally done away with by adopt-
ing a barrel - shaped furnace, shown in Fig. 15, which
revolves around its long axis. The furnace proper a con-
sists of an iron shell lined inside with firebricks. The
shape is that of a barrel; it either conforms to the

62 ALKALIES AND HYDROCHLORIC ACID § 29

«o

' T I T T

r—T

T-r

outside shell, or, if that is cylindri-
cal, the bricks are laid thicker at
the ends than in the middle. Two
rows of these lining bricks are laid
higher than the rest, to break up
the mass and mix it, and also to
better expose it to the fire gases
as it drips from these projections.
These furnaces are from 15 to
30 feet long and average about
6 feet in diameter at the ends, and
from 10 to 12i feet in the middle;
they are heated by the fire gases,
which pass in at one end and out
at the other. The furnace is
heated by the gases from the
grate at r, or sometimes by pro-
ducer gas, although for some
reason this latter does not seem
to be much used. The hot gases
pass into a, where they bring
about the conversion of the salt
cake into black ash, and then pass
out through e to ^, where they
pass over the top of pans contain-
ing the liquor from the lixiviation
of black ash and evaporate it.
At b is shown the manhole
through which the black-ash mix-
ture is introduced, and from
which, at the end of the process,
the black ash is discharged into
the wagons d: The draining pan
for^the black salts is shown at A.

61. Chargre for the Meclian-
Ical Fiimace. — The theoretical
charge for the mechanical furnace

§ 29 ALKALIES AND HYDROCHLORIC ACID 53

will naturally be the same as already calculated for the hand
furnace, and. the same conditions of water and impurities
in the limestone and coal rule here. It is found, however,
that there is less burning of the mixing coal and less
mechanical loss of the constituents of the mixture, so that
not so large an excess over that theoretically demanded is
now used for the mechanical furnace. The average propor-
tions of the constituents of the black-ash mixtures to be used
with a mechanical furnace are 100 parts of salt cake, 82 parts
of limestone, and 30 parts of good coal. The size of the
charge will naturally vary with the size of the furnace, but
an average charge is from li^ to 3 tons of the mixture.

68i Managrement of tlie Mechanical Furnace. — The

operation consists in charging in all the limestone and
about two-thirds of the coal, without drying. The cover is
then put on and the cylinders slowly revolved (about 1 revo-
lution in 3 to 4 minutes) until the appearance of a bluish
flame of carbon monoxide around the manhole shows that at
least a part of the limestone has been converted into lime.
As soon as this operation is completed, which requires from
'1 to IJ hours, the cylinder is turned so that the charging
hole is up and the finely ground salt cake and the coal are
dumped in. The cover is then replaced, the draft through
the cylinder diminished, and the slow turning resumed.
After about 15 minutes, the mixture is hot enough so that
the danger of carrying away parts of the mixture is not so
great and the draft is restored; in a few minutes, the
appearance of a bright yellow flame around the manhole
shows that a part of the charge is becoming fused. The
rate of revolution of the cylinder is then brought up to 3 or
4 revolutions per minute. The charge is now watched
through peep holes and when yellow flames (candles) are
seen to break from it, it is time to stop. The furnace is
now revolved a few times as quickly as possible to bring the
mass together. It is then turned so that the charging hole
is up, the cover is removed, and the furnace turned so that
the charge runs out into the wagons d. If the furnace is

54 ALKALIES AND HYDROCHLORIC ACID § 29

worked properly, the gas should continue to be given off
while the material is in the barrows and thus a porous black
ash is produced.

In some works the method proposed and patented by
Mactear is adopted. This consists in making a mixture of,
say, 100 parts of salt cake, 73 parts of limestone, and 40 parts
of coal. This mixture is put together into the furnace and
the reaction brought to an end, as shown by candles, then
from 6 to 10 per cent, of the weight of salt cake, of quick-
lime, and from 14 to 16 per cent, of furnace cinders are
added and the furnace turned quickly two or three times to
thoroughly mix the materials and then the whole run out.
This method saves considerable time in working, as the
preliminary conversion of a portion of the limestone into
lime is saved, and the material is left in a condition con-
sidered by many to be the best for lixiviation.

63. Advantagres and Disadvantagres of the Mechan-
ical Furnace. — The mechanical furnace has the advantage
over the hand furnace that it makes the manufacturer more
independent, as the only skilled man needed is the foreman,
and he can tend to several of these furnaces. It gives a
large output with a comparatively small amount of manual
labor, and at the same time a more uniform material is
obtained.

On the other hand, the revolving furnaces are expensive
to build, and as frequent repairs are necessary, they are
expensive to maintain.

64. Cyanides. — One of the most disagreeable impuri-
ties occurring in black ash is the sodium cyanide formed
from the nitrogen in the coal. This cyanide will unite with
iron, if opportunity is offered, and make sodium ferrocyanide,
which it is very hard to remove from the solution, but which
decomposes at the end, when the soda ash is calcined, into
sodium carbonate and ferric oxide, coloring the soda
ash. In the hand furnaces, usually no attempt is made to
remove the sodium cyanide from the black ash, but for
mechanical furnaces the Pechiney-Weldon method works

§ 29 ALKALIES AND HYDROCHLORIC ACID 55

nicely. This process depends on the fact that when sodium
cyanide is fused with sodium sulphate the cyanide is decom-
posed. It is not known exactly what the reaction is, but
probably the following equation very nearly expresses the
truth :

Na^SO, + %NaCN= Na^S + Na^CO, + C(9 + iV,

The operation consists in adding a little salt cake to the
first finished black ash in the furnace, giving the furnace a
few turns to mix the charge thoroughly, and then dis-
charging the black ash at once into the barrows. The
amount of salt cake required must be determined for
each furnace and mixing coal, as the amount of cyanide
will vary as these conditions vary. As there is no time to
analyze the black ash just before adding the salt cake, a
fixed amount must be decided upon and then added to each
charge of ash. This is best done by determining the
amount of cyanides in several charges of black ash from a
furnace, averaging these, calculating the amount of salt
cake necessary by the above equation, and then adding from
four to six times the theoretical amount to the charge each
time just before emptying the furnace, as above stated. For
example, if an average analysis shows ^ per cent, of sodium
cyanide, there will be ^ pound of sodium cyanide in
100 pounds of the mixture, and from the equation

Na^SO, + %NaCN= Na^S + Na^CO, + CO + N^

we have 142 : 98 = ;ir : .5; or, theoretically, it will require
.72 pound of sodium sulphate. It is not very easy to get
material of this character in extremely close contact, how-
ever, so the excess is necessary, and if we select six times
the theoretical amount, we should add 4.32 pounds of salt
cake for every 100 pounds of the mixture used. This is
rather an extreme case, as usually the cyanide will not run
so high.

Another method for attaining the same end, and one that
is much preferred by many manufacturers, consists in add-
ing regulated amounts of salt cake to each furnace charge

56 ALKALIES AND HYDROCHLORIC ACID § 29

until the amount is found that gives the most satisfactory
result.

So far nothing has been said concerning the excess of
salt cake added, and we might naturally consider that there
would be excess of lime and coal enough in the black ash to
convert it into sodium carbonate, and no doubt there is. It
has been found better practice, however, to add about an
equal weight of finely ground limestone to the salt cake
used before adding it to the mixture in the furnace.

A mixture that has given good results with this process is
salt cake, 100 parts; limestone, 78 parts; coal, 37.5 parts;
and as a final addition, a mixture of 6 parts of salt cake and
7 parts of powdered limestone.

65. Properties of Black Ash. — A good black ash from
a hand furnace should have on the fracture a brownish-
black or dark slate-gray color, and a porous, pumice-like
structure. It should be uniform in appearance throughout
the ball arid should not have many black spots of coal or
white ones of limestone. Balls that are pale pink or reddish
are usually also dense and burned, and will be found on
analysis to be high in sodium sulphide and sodium sulphate.
Each man's work for the day should be tested in the labora-
tory for, at least, total alkali, sodium sulphide, and sodium
sulphate.

Black ash from a mechanical furnace appears quite differ-
ent from that from a hand furnace, being dense and of a
higher color. It would be almost impossible to lixiviate this
ash were it not for the free limestone contained in it, which,
on slaking, breaks up the pieces of black ash, so that the
water can get at it to dissolve out the sodium carbonate.

66. Composition of Black Ash.. — Black ash naturally
varies somewhat in composition, but usually has about 40 per
cent, of soluble matter, consisting of the carbonate, oxide,
chloride, sulphate, sulphite, thiosulphate, aluminate, silicate,
cyanide, and sulphocyanide of sodium ; while the insoluble
portion consists mainly of the sulphide, carbonate and oxide

§ 29 ALKALIES AND HYDROCHLORIC ACID 57

of calcium, ferrous sulphide, aluminum oxide, silica, mag-
nesium oxide, carbon, sand, and insoluble sodium compounds
of aluminum and silicon. Of course, sodium carbonate, cal-
cium sulphide, and calcium oxide are the preponderating
substances.

67. lilxlvlatlon of Black Ash. — The black ash when
removed from the furnace is very hot and must be allowed
to lie and cool until it can be conveniently broken and
handled. This usually requires about 2 days. It should
not, however, be allowed to lie longer than is necessary, for
the moisture, carbon dioxide, and oxygen of the air act upon
it. The carbon dioxide converts the lime into calcium car-
bonate, and the calcium sulphide into calcium carbonate
and hydrogen sulphide. The oxygen converts calcium sul-
phide to calcium sulphate and various intermediate oxida-
tion products. Finally the moisture aids in the formation
of sodium sulphate, sulphide, etc., from the calcium salts
and sodium carbonates, and thus causes a loss of the valuable
sodium carbonate.

Various difficulties must be overcome in the lixiviation ;
for the lime is slaked and tends to react with the sodium
carbonate, as above, while the calcium sulphide also reacts,
to form the sulphide of sodium. This takes place especially
rapidly if the solution is hot and dilute. Furthermore, the
oxidation of the calcium sulphide to sulphate and then a
reaction between that and the sodium carbonate takes
place here as well as in the preceding case, unless the mate-
rial is protected from the air. It is necessary then to lixivi-
ate away from the air as rapidly as possible and to keep the
liquid cold. These last two conditions seem to be and are
directly opposed to each other, but the temperature is
selected that will give the most rapid extraction with the
least trouble in other directions.

68. Shank^s lilxlvlatlon System. — This system for
lixiviating the black ash has practically displaced all other
systems, as it is rational, simple, and efficient. The lixivi-
ating apparatus consists of one large tank divided into from

68 ALKALIES AND HYDROCHLORIC ACID § 29

four to eight water-tight compartments. Each compart-
ment has a false bottom of perforated sheet iron, which
serves to support the lumps of black ash and acts as a filter
for the solution of sodium carbonate. A pipe leads from
under the false bottom of each section of the apparatus to
near the top of the other sections, so that the different sec-
tions may be connected together at will; each section has
a pipe for supplying fresh water when necessary. Fre-
quently they are fitted with steam connections as well, so
that the liquid may be warmed, if desirable.

In working, the water or dilute lye flows in at the top of
the section, and as it dissolves more material it becomes
heavier and sinks to the bottom of the tank ; it is then forced
into the next tank by the fresh incoming lye ; this process is
continued until it finally flows away sufficiently concentrated.
The pipes are so arranged that the contents of the various
tanks are always completely covered. The process is con-
tinuous, the water flowing into the tank containing the most
nearly extracted black ash and flowing away from the last
and most recently charged tank as long as the specific
gravity does not fall below 1.25. As soon as the specific
gravity of the lye from the last tank falls below 1.26 it is
turned into a tank recently filled with new ash and the
exhausted ash is washed with water until the wash watei
has a specific gravity of only 1.005. Then the waste is sent
to the dump and the tank is freshly charged to serve as the
end tank in its turn. The best temperature for lixiviation
to give concentrated solutions is about 50° C, which is
usually reached by the heat from the slaking of the lime
when the lye comes in contact with the fresh ash. If this
does not occur, the temperature can be raised by blowing in
steam. In the first one or two tanks of the series, where the
lye is weak, the temperature is not allowed to get below
35° C. The black ash, as can be shown by extracting with
alcohol, contains no sodium hydrate or sodium sulphide, the
lye obtained from its lixiviation contains not only these sub-
stances, but various other soda compounds formed by inter-
change during the lixiviation.

§ 29 ALKALIES AND HYDROCHLORIC ACID 59

Although the composition of the various lyes differs con-
siderably, depending on the conditions of lixiviation, etc.,
the following analysis of a lye of 1.25 sp. gr. will give an
idea of the general character of such solutions. The solu-
tion contained 313.9 grams of solid substance per liter, and
the solid had the following composition :

Sodium carbonate . . 71.30$^ Sodium sulphate 24j^

Sodium hydrate 24.505^ Sodium cyanide 09j^

Sodium chloride. . . . 1.90$^ Alumina 1.51^^

Sodium sulphide 10^ Silica 19^

Sodium thiosulphate .37^ Iron traces

69. Purlflcation of the Tjye, — The lye contains con-
siderable finely divided suspended matter, and is therefore
allowed to stand for a time in a warm place to allow it to
settle and become clear. The iron compounds if left in the
lye would decompose at a later stage of the process and
color the ash. The sodium ferrocyanide may be decomposed
by heating the lye to 180° C, the following reaction taking
place between the sodium ferrocyanide, sodium thiosul-
phate, and sodium carbonate:

+ bNa^SO, + NaCHO^ + NH^ + ^NaHCO, + FeO

This method is, however, difficult and expensive, so that
it is far better to use the Pechiney-Weldon method and so
exclude the ferrocyanide from the black ash, and thus from
the lye. The iron sulphide may be separated by allowing
the lye to stand exposed to the air, when the iron sulphide
slowly separates out. This, method is slow, hgwever, and it
is better and more usual to allow the lye to flow down ropes
and chains in tall towers, up which are passing carbon diox-
ide and oxygen from the black-ash furnaces or from lime
kilns. By this means the caustic soda is carbonated, form-
ing sodium carbonate; the iron is precipitated, and the
sodium sulphide is converted into sodium carbonate with the
liberation of the hydrogen sulphide. This last reaction is
not complete under practical conditions, so that sometimes

60 ALKALIES AND HYDROCHLORIC ACID § 29

zinc hydrate is mixed with lye at this point to complete
the removal of the sodium sulphide.

70. Pauirs Method. — This method for purifying the
tank liquor consists in mixing it with a little Weldon mud
(see Alkalies and Hydrochloric Acid^ Part 2) and then blow-
ing in air and steam until the sodium sulphide is thoroughly
oxidized and the iron, silica, and alumina are precipitated ;
about 2 pounds of manganese dioxide to every 100 pounds
of sodium carbonate in the solution is a suitable proportion,
although sometimes a smaller amount of the manganese
dioxide will work very well.

If, for the sake of convenience, we consider Weldon mud
as manganese dioxide, the reactions may be written as
follows:

%Na^S + ^U^nO^ + hHfi = %NaOH^ Na^^fi^ + 4:Mn{0H)^

Ufn{0//), + 2(?, = iMnO^ + ^Hfi

Since the manganese dioxide is continuously recovered,
except the small amount carried away mechanically, it may
be used over and over until, through the precipitation
of ferric hydrate, silica, aluminum hydrate, etc., the pre-
cipitate becomes too bulky to handle, when it must be
thrown out and new Weldon mud supplied.

71. Evaporation of the Tank Liquor, — The tank
liquor after settling and purification is evaporated to obtain
the sodium carbonate. We may conveniently divide the
methods for evaporating the tank liquor into three classes,
i. e., in pans by surface heat, in pans by heat underneath,
and in pans with mechanical stirrers, by means of which the
sodium carbonate crystals are fished out as soon as formed.
Of these three methods, that using surface heat is the most
common ; it is very convenient, for it utilizes the waste heat
from the black-ash furnace.

78. Surface-Heat "Evaporation. — The pans for this
purpose are shown in connection with the black-ash furnaces
in Figs. 14 and 15. They are of very simple construction
and are made of about | -inch sheet iron. They are provided

§ 29 ALKALIES AND HYDROCHLORIC ACID 61

with two or three doors, as the case may be, and are so
formed that the contents (crystals and mother liquor) can
be drawn out on the draining table h in Figs. 14 and 15.
During the evaporation of the liquor the doors are closed,
and to make them tight, are luted on with clay.

In working the pan after the doors are closed, the pan is
filled with the clear settled liquor and the waste gases from
the black-ash furnace allowed to pass over the surface of the
liquor. This soon brings the liquor to a boil, and the cur-
r-ent of hot gas, by carrying away the vapor as fast as it is
formed, rapidly concentrates the solution. From time to
time fresh liquor is run in until the pan is nearly filled with
crystals, when the evaporation is allowed to continue until
the mixture of crystals and mother liquor has about the con-
sistency of mortar. The doors are then removed, the mother
liquor allowed to run off, and the whole mass brought on to
the draining table. The mother liquor, **red liquor," is
allowed to drain off until another panful is nearly ready to
run out, when the crystals are removed to a special drainer,
where they are allowed to lie and drain 24 hours.

The surface evaporation has the advantage that it is rapid,
but the disadvantage that the sulphur dioxide from the fire
gases is all absorbed here and causes a loss of sodium carbon-
ate. Dust from the black-ash furnace is also carried over
into the pan and makes the salts impure.

73. Pans With. Heat Below. — Pans heated below have
the disadvantage that they do not last so long and that they
are neither so effective nor economical, but, on the other
hand, they give a purer product and the loss of sodium car-
bonate, through the acids in the heating gas, is avoided.
Various shapes of pans are in use for this purpose, but those
built boat shaped (i. e., with sloping sides and narrow bot-
tom) and heated more along the sides than on the direct
bottom, are the best; for in these, by the boiling, the sodium
carbonate crystals as they separate settle in the narrow,
bottom portion of the pan, where they are away from the
direct heat of the fire and from which place they can be

62 ALKALIES AND HYDROCHLORIC ACID § 39

scooped out. There is always more or less trouble even
with this style of pan, however, through the crystals burn-
ing fast to the bottom of the pan.

74. Mechanical Pans. — These pans are also heated by
outside fire, but they have mechanical stirring devices that
not only prevent the crystals sticking to the bottom of the
pan, but save labor by working the crystals to the end of
the pan and finally lifting them out to drain. By this sys-
tem, fresh liquor can be run in continuously and the salts
removed until the mother liquor gets too thick with caustic
soda and sodium sulphide, when it is drawn off and fresh
liquor started again. The most satisfactory pan of this

type is the Thelan pan shown in Fig. 16. This consists of
a semicircular iron pan ti, which is heated on the outside by
the fire from the grate i/. The hot gases circulate under
the pan and escape to the chimney at the opposite end. The
scrapers d, which are rotated by the shaft and gear c, prevent
the separated salt from burning fast to the pan and move it
to the end, where it is lifted to a draining apron. From the
draining apron the salt is moved to a large draining table,
where it is allowed to drain H hours before being calcined.

§ -29 ALKALIES AND HYDROCHLORIC ACID 63

75. CalelnlnflT the Crystals. — The salt that separates
in the evaporating pans is dark in color and is known as
the black salt. It consists mainly of monohydrated sodium
carbonate Na^CO^^Hfi^ and must be calcined to remove
the water and oxidize any remaining sodium sulphide and
organic matter. The calcining usually takes place in a
reverberatory furnace similar to a black-ash furnace, and
the charge may be brought to a dull-red heat, but must not
be fused. During the drying, the material must be turned
over occasionally and the lumps broken up, but further than
this the operation requires very little attention, outside of
the charging and discharging of the furnace and tending
to the fire.

76. Grinding^ the Soda Ash. — By calcining the black
salt, the material is caused to cake together so that it is
necessary to g^ind it before putting it on the market. This
operation is carried out in ordinary mills, such as are used
in grinding grain in making flour.

SODA CRYSTALS

77. Sodium carbonate crystallizes at ordinary tempera-
tures with 10 molecules of water, forming crystals gen-
erally known as sal soda, or washing soda. These crystals
contain 63 per cent, of water, and many people consider the
crystallized material so much better than the calcined soda
ash that they are willing to pay the freight on all the water
in order to have the crystals. This attitude was justined
before ammonia soda came into the market in such large
quantities, for the soda crystals were purer than any of the
soda ash then available. At the present time most of the
crystal soda is sold for household purposes. It is better
than soda ash for laundry purposes, for it dissolves quickly,
and so avoids the danger of particles of the undissolved
soda getting on the linen and damaging it. The soda
crystals NajOO^^\^Hfi are manufactured from the cal-
cined soda ash. This substance is dissolved in hot water

64 ALKALIES AND HYDROCHLORIC ACID § 29

and allowed to stand and settle until quite clear, when it
is run into iron crystallizing pans. The size and shape of
these pans vary considerably, but these features are not of
material importance; the essential thing is a pan that will
cool slowly and not render the solution impure. These pans
are allowed to stand from 5 days, in winter, to 16 days, in
summer, for all the crystals that will to separate. When the
crystallization is seen to be complete, by no more crystals
forming, a hole is broken in the crust and the mother liquor
drawn off. The crystals are then drained and packed.

Soda crystals made from pure soda ash are soft and
unsatisfactory, so that it has been found advisable to have
enough sodium sulphate in the solution that the crystals will
contain from 1 to IJ^ per cent, of sodium sulphate. For
some reason, this admixture of sodium sulphate renders the
crystals hard.

78. Yield. — Owing to a number of causes, only about
70 per cent, of the sodium occurring in the sodium sul-
phate is finally obtained as sodium carbonate. The main
sources of loss are a mechanical carrying away of part of
the charge by the fire gases in the black-ash furnace, and
a volatilization of another part by the high heat. A
portion of the sodium sulphate fluxes, with the brick lining
of the furnace and the coal ashes, and forms insoluble
sodium compounds. There is always a more or less incom-
plete conversion of sodium sulphate into sodium carbonate,
and a further loss by, necessarily, incomplete lixiviation.
Finally, the action of the water in causing a reverse reaction
causes a loss of soda.

79. Finished Soda Ash. — The finished product from
the Le Blanc method of the manufacture of soda ash should
be nearly, or quite, white and should show very few reddish
specks after grinding. It should not contain over 2^ per cent,
of sodium hydrate (unless intended for special purposes), nor
should the insoluble matter exceed 1 per cent. It should not
be possible to detect sulphides in it, and the sulphite should

§ 29 ALKALIES AND HYDROCHLORIC ACID 65

not exceed .1 per cent. Sodium chloride and sulphate are
always present and are harmless, but they should not exceed
4 per cent.

80. Uses of Sodium Carbonate. — Sodium carbonate
is used for an almost unlimited number of purposes, for
some of which sodium bicarbonate, or caustic soda, is also
used and frequently to better advantage than when soda ash
is employed. The most important uses for soda ash may be
enumerated as follows:

(1) The manufacture of the various kinds of glass. In
the place of soda, salt cake is frequently used for this
purpose. (2) The making of various kinds of hard soap.
Caustic soda is also used for soap making. (3) The manu-
facture of borax and various other sodium compounds.
(4) In the preparation of starch, the manufacture of glu-
cose, the preparation of the fatty acids, the purification of
oils and of pyroligneous products, and otherwise in the
organic manufactures. (5) For scouring, dyeing, etc. in
cloth manufacturing.

81. Methods for Stating: Stren^h of Soda Ash.

Soda ash may contain varying amounts of sodium sulphate,
sodium chloride, and various other substances that have no
value as alkali. The methods of determining the amount
of available alkali in a sample of soda are more suitably
explained in a treatise on chemical analysis; but since the
methods for stating this value vary considerably, it is desir-
able that they should be explained here.

The French express the value of their soda ash in
degrees Descroizilles. It is based upon the reaction
between sodium carbonate and sulphuric acid, and is
expressed in terms of the number of parts, by weight, of
100 per cent, of sulphuric acid that are necessary to neu-
tralize 100 parts of the substance. Since 53 parts, by
weight, of sodium carbonate neutralize 49 parts, by weight,
of sulphuric acid, then 100 parts, by weight, of chemically
pure sodium carbonate will neutralize 92.45 parts, by

ALKALIES AND HYDROCHLORIC ACID § 29

weight, of sulphuric acid ; therefore, chemically pure soda is
92.45° Descroizilles. By the same reasoning, chemically
pure sodium hydroxide is 122.5° Descroizilles.

The Germans very rationally report the percentage of
sodium carbonate in the sample. Since, however, by the
method of determining this percentage, caustic soda will also
be determined and reported as carbonate, which may have
the peculiar effect of showing a substance to be 120 per
cent, pure, this method is not so suitable as the English
method.

TABIiE II

Percentage
Sodium

Actual
Alkali
Na^O

English
Alkali

Liverpool
Alkali

Descroizilles

Carbonate

Test
Na^O

Test
NaaO

Degrees

795^

46.5

47.11

48.00

73-57

82.07

48.0

48.63

49 -M

7587

85.48

50.0

50.66

51.61

79 03

88.90

52.0

52.68

53-67

82. 19

90.61

530

53-70

54-70

83-77

94 03

55-0

55-72

56.77

86.93

97-45

57.0

57.75

58.83

90.09

99. 16

58.0

58.76

59.87

91.68

100.02

58-5

59-27

60.38

92.45

The English rate their alkali on the percentage of real
or available alkali; that is, on the percentage of Nafi in
the case of both sodium carbonate and hydrate. This
method seems to be the most sensible, for it is the real
alkali that is of value, and it does not matter so much in
what form it is; therefore, the percentage of the valuable
constituent is given. This system is also somewhat used
in France and is there called the Gay-Lussac degree.
Unfortunately, when this system was established in Eng-
land, the values of the atomic weights were not exactly

§ 29 ALKALIES AND HYDROCHLORIC ACID 67

determined, and so 32 was used as the equivalent weight of
Nafiy instead of the more correct value 31. Although it is
now well known that this error exists, it is still retained,
either through dishonesty or a neglect to change. Besides,
in the Liverpool district, a mistake was, and is, made in such
a way in making the calculation that an even greater error
is made in the manufacturers' favor in stating the strength
of the alkali.

In the United States, the English system is pretty gen-
erally adopted, using the correct equivalent for Nafi\
although in New York and some other large cities, where
considerable soda is imported from England, the English
and Liverpool degrees are also in use.

Table II shows the relation between the different meth-
ods for stating the value of soda ash.

TANK WASTE

82. The residue that is left after the removal of the sol-
uble constituents from the black ash consists mainly of the
sulphide and carbonate of calcium with small amounts of
various other substances and is generally called the tank
waste. Practically all the sulphur that was contained in
the sodium sulphate is left in this waste, and therefore,
unless it can be recovered, it represents an enormous loss
of money. In addition to that, it requires room for dumps,
and by weathering it produces an almost intolerable nui-
sance, due to the escape of hydrogen sulphide and sulphur
dioxide into the air. The weathering of the tank waste
also causes the formation of polysulphides of sodium and
calcium, forming the so-called yellow liquors, which run
into the streams and sewers and contaminate them and
which also saturate the soil of the neighborhood, spoiling
the wells and doing other damage.

Table III gives an idea of the composition of tank waste
from the mechanical furnace and the hand furnace.

68 ALKALIES AND HYDROCHLORIC ACID § 29

TABIiE III

COMPOSITION OP TANK WASTE FROM MECHANICAX

AND HAND FURNACES

Constituents

Sodium carbonate . . .
Calcium carbonate . .
Calcium hydrate ....

Calcium sulphide

Calcium thiosulphate

Calcium sulphite

Calcium sulphate . . .

Calcium silicate

Carbon

Alumina

Ferrous sulphide. . . .
Sand

Revolver.
Per Cent.

2.9
24.7

i.o

54.7

.5
trace

trace

2.5
8.4

1-5

2.0

Hand Furnace.
Per Cent.

2.5

33-2
9.0

37-3
2.0

1.0
6.4

•5
2.5
5-0

These analyses are made on the dry substance, so that in
addition to the above we must calculate about 30 per cent,
of water in the composition of the waste.

The disposal of this waste material has been one of the
important problems of the Le Blanc manufacturer ever since
the industry became of sufficient importance for the waste
to be noticed, and it still continues to trouble him, although
the problem has been fairly well solved. It is best disposed
of, when the works are located near the coast, by loading it
upon scows, towing it out to sea, and dumping it. This,
of course, wastes the sulphur, but it avoids the nuisance.
Where it cannot be conveniently sent to sea and it does not
pay to employ one of the recovery processes to work it up,
the waste is spread out evenly and then packed down to
prevent, as far as possible, the infiltration of rain.

The processes that have been proposed for recovering the
sulphur from the waste are numerous, but only one has been

§ 29 ALKALIES AND HYDROCHLORIC ACID 69

permanently successful. Even now only a part of the waste
is worked for sulphur recovery.

83. Chance-Claus Process. — The only process that
has ever been commercially successful and the only one
that is in successful operation today for the recovery of the
sulphur from tank waste, is the so-called Chance-Claus proc-
ess. This process depends essentially on the decomposition
of the waste by carbon dioxide, which reaction was proposed
by Gossage in 1836. He believed in the process so thoroughly
that he spent 30 years of his life and a fortune in money
striving to perfect it, but without success. His principal
difficulty was that he could not get the escaping gas rich
enough in hydrogen sulphide and its composition varied too
much. The attainment of this result, together with a
method for getting the sulphur from the hydrogen sulphide,
comprise the achievements of Chance and Claus in this
direction.

84. In carrying out the process, the tank waste is made
to a slurry with water and then charged into a cylinder. A
battery of seven cylinders is usually employed, which are so
arranged that the gas can be passed from one cylinder to
any other. In operation six cylinders are in use and one is
being emptied and recharged. The gas used must be of
regular composition and contain not less than 30 per cent,
of carbon dioxide. This is best obtained from lime kilns
similar to those used in the ammonia-soda process. It
is passed into the cylinder containing the most nearly
exhausted material and sets free the hydrogen sulphide
according to the reaction

Ca{SH)^ + CO, + H^O = CaCO, -f 2//,5

This hydrogen sulphide passes into the following cylinders,
where it is absorbed by the calcium sulphide

CaS+H,S^Ca{SH\

Since the most recently charged cylinder is placed last,
the hydrogen sulphide is practically all absorbed and the
escaping gas is almost completely free from it and might

70 ALKALIES AND HYDROCHLORIC ACID § it9

escape directly into the air. For the sake of safety, how-
ever, it is usually run through a purifier similar to those
used to purify coal gas and containing either oxide of iron
or lime. When the contents of the last two or three cylin-
ders are nearly converted into calcium sulphydrate, the
escaping gas begins to be stronger in hydrogen sulphide.
At this point the back cylinders are tested to see if the gas
will burn, for this is an indication that it is 30 per cent., or
stronger, in hydrogen sulphide. As soon as the gas from
one of the intermediate cylinders is found to be strong
enough, it is put in connection with a gas holder and the gas
collected until its composition falls below 30 per cent, of
hydrogen sulphide. (The water lute of the gasometer is
shut off from the air by a heavy layer of oil to prevent the
escape of the gas into the air.) When the gas contains less
than 30 per cent, of hydrogen sulphide, it is turned into
freshly charged cylinders, and the first cylinder, the con-
tents of which should be so free from sulphides by this time
that they do not blacken lead paper, is emptied and
recharged with fresh slurry. The water from this residue
is so pure that it can be run directly into the streams and
the solid material, which contains over 85 per cent, of
calcium carbonate, can be used for fresh black-ash mix, or
for making cement.

85. The hydrogen sulphide is so strong that it can be
burned direct for the manufacture of sulphuric acid, and it
yields an exceedingly pure acid free from arsenic. The
greater part of the gas is converted into sulphur, however,
for the sulphur is more valuable in the free condition than
in sulphuric acid. The thing that has probably done the
most to make the Chance-Claus sulphur-recovery process
commercially successful is the method of converting the
hydrogen-sulphide gas into sulphur. This consists in pass-
ing through iron oxide heated to dull redness a mixture of
hydrogen sulphide and air in the proportions given by the
equation

g 29 ALKALIES AND HYDROCHLORIC ACID 71

When the kiln is tirst started, it is necessary to heat the
iron oxide to the proper temperatnre; but when once
started, the reaction keeps the temperature of the oxide
high enough to continue the reaction.

86. ClauB Kiln. — Fig. 17 shows the Claus kiln as at
present used in the Cbance-Claus sulphur-recovery process.

72 ALKALIES AND HYDROCHLORIC ACID § 29

The gas is mixed in the gasholders with a proper amount of
air for its decomposition, according to the preceding equation.
The composition of this mixture must be very carefully
determined by analyses, and the amount of air regulated so
that there will be just sufficient oxygen to burn the hydro-
gen of the hydrogen sulphide, but no excess. By deter-
mining the amount of hydrogen sulphide in the gas in the
holder, it is easy to calculate the amount of air necessary to
add to make the proper mixture. From the equation it is
seen at once that each volume of hydrogen sulphide requires
^ volume of oxygen. Then, if the gas in the holder is
32 per cent., by volume, hydrogen* sulphide, each liter of
the gas will contain .32 liter of hydrogen sulphide, which
will require .16 liter of oxygen, but the air only contains
21 per cent, of oxygen, so that we must take ^f liter of air,
or .76 liter of air. That is, 3 volumes of air must be mixed
with every 4 volumes of the gas from the holder. Of course,
when the gas from the holder has a different composition,
the amount of air must be varied; so that it is very essential
for the success of this process that the gas be of a very uni-
form composition, and that the work be constantly controlled
by analyses. The gas mixture passes from the gas holder
through its conduction pipe and the lute A, to prevent the
flame from striking back and exploding the gas holder, into
the top of the kiln proper B, This is made of iron and is
about 9 feet high and, on an average, 25 feet in diameter; it
has a grate that bears a layer of broken bricks, on which is
about 12 inches of ferric oxide. At first the gas was passed
in at the bottom of the kiln, but it was found that here, as
is generally the case where a gas must come in ' intimate
contact with a solid, a better result is obtained by passing
the gas mixture down through the oxide. In starting a
kiln, a fire is built on the iron oxide and kept going until the
oxide is red hot; the gas mixture is then turned in and the
reaction between the oxygen of the air and the hydrogen
sulphide takes place. The temperature of the oxide is kept
up without any further outside heat. The best temperature
is about 230° C, taken at the exit pipe from the kiln. The

§ 29 ALKALIES AND HYDROCHLORIC ACID 73

reaction is a reversible one, so that it will never be quite
complete, and it is not possible to add an excess of oxygen
to force it, for in that case sulphur dioxide in too large
quantities would be formed.

From the kiln the products pass into a small chamber C,
where the molten sulphur deposits, while the gases and sul-
phur vapor pass into the larger chamber D^ where the
sulphur vapor deposits as flowers of sulphur and some of the
steam is condensed. This chamber contains walls part way
across, as shown in the figure. These walls serve as baffle
plates and separate the fine sulphur, which would otherwise
be carried into the washing tower and clog it, and at the
same time be lost. From D the gases pass through the
washing tower -£", down which water is kept flowing to
remove sulphur dioxide from the gas. It then passes
through a purifier /^containing lime or iron oxide to remove
the last of the hydrogen sulphide, so that the gas escaping
into the air is practically pure nitrogen.

This process, when working well, recovers from 85 to
90 per cent, of the sulphur in the waste and entirely abates
the nuisance otherwise due to the waste decomposing in the
open air. The cost of the installation of the plant is small
and its operation is not expensive, but the price of sulphur
is at present so low that it hardly pays to recover it, and if
it were not for disposing of the waste, the process would
probably go out of use.

87. Sodium Tlilosnlphate. — Since sodium thiosulphate,
or what is more commonly known as sodium hyposulphite,
or hypo, is made almost exclusively from tank waste, it
deserves a few words here. It is made by blowing air
through the waste suspended in water until all the sulphide
is converted into calcium sulphite and thiosulphate, and then
adding sodium sulphate, or carbonate, which gives the insol-
uble calcium salt and leaves sodium thiosulphate in solution.
This is boiled with sulphur to convert the sulphite into
thiosulphate, then crystallized out and purified by recrystal-
lization.

74 ALKALIES AND HYDROCHLORIC ACID § 29

Another method is to pass sulphur dioxide intie-the waste,
thus converting the sulphide into thiosulphate according to
the reaction

and then converting it into the sodium salt as above.
Sodium thiosulphate forms soluble salts with silver, thus
dissolving silver iodide and chloride. For this reason it is
largely used in photography and the metallurgy of silver
It is also used as an antichlor in paper making, in certair
kinds of dyeing, and for various other uses.

SODIUM HYDRATE

88. Historical. — The manufacture of sodium hydrate
on the large scale at a factory does not date back nearly so
far as the manufacture of soda ash. It is true that caustic
soda has been used for soap making almost as long as soap
has been known, but for a long time it was made at the soap
manufactory and used in the form of solution. It was not
until 1850 that the manufacture of caustic soda, as such,
began, and then only on a small scale ; and it was not until
1860 that the manufacture attained any considerable impor-
tance. From that time on, however, more and more caustic
soda has been made, until now it is an important branch
of the alkali industry.

89. Sodium Carbonate and Lilme. — The most common
process for the preparation of caustic soda is by means of
the reaction between sodium carbonate and slaked lime.
This reaction is

Na^CO, + Ca{OH)^ = 2NaOH+ CaCO^

•

Since the reaction is a reversible one, it is not desirable to
make the sodium hydrate too strong, for the stronger the
solution is in caustic soda, just so much more tendency is
there for it to go towards the formation of calcium hydrate
and sodium carbonate. On the other hand, although dilute
solutions lead to a high percentage transformation of the

§ 29 ALKALIES AND HYDROCHLORIC ACID 76

sodium carbonate, they require large apparatus and much
heat to drive off the water, in the making of the solid caustic.
It is, therefore, necessary to pursue a middle course. A
solution of sodium carbonate of 1.1 sp. gr., that is, about
10 per cent., is generally considered to be the most advanta-
geous strength for conversion into the hydrate. With a
solution of this strength, about 97 per cent, of the sodium
carbonate used can be converted into caustic soda, which
gives a fair strength of solution.

CBITDB MATEBIAIiS

90. Soda Ash.. — The soda from the Le Blanc process is
well suited for making caustic soda, for it frequently contains
considerable caustic, which has been formed by the lixiviation
of the black ash, and so requires less lime than would other-
wise be the case. By the addition of a large excess of lime-
stone to the black-ash charge, practically all the sodium can
be obtained as the hydrate; this method is sometimes
employed. A suitable furnace charge to employ when the
tank liquor is to be used for making caustic is 100 parts,
by weight, of salt cake, 110 parts of limestone, and 65 parts
of coal. A part of the limestone is frequently replaced by
caustic mud (see Art. 97) in the proportion of about 20 parts
of the mud to 12 parts of limestone. The red liquid (mother
liquor from the black salt) from the Le Blanc process, in
which is concentrated much of the caustic originally in the
black ash, is frequently utilized for making caustic soda.

At the ammonia-soda works the sodium bicarbonate mixed
with water is first boiled by steam in a closed apparatus, so
that the ammonia and from 75 to 80 per cent, of the bicarbo-
nate carbon dioxide are driven off and utilized in the car-
bonating towers, while the sodium-carbonate solution, which
contains about 20 per cent, of the bicarbonate, is used for
making caustic soda.

91. Inline. — The lime used for making caustic must be
of good quality, for a low percentage of CaO not only makes

76 ALKALIES AND HYDROCHLORIC ACID § 29

necessary the introduction of large amounts of impurity into
the causticizing tank, but it also gives a caustic liquor that
settles badly and so interferes with the work. A satisfactory
lime should contain at least 85 per cent of CaO.

DETAIT^S OF THE PROCESS

92. Caustlclzlngr the Sodium Carbonate. — The caus-
ticizing of the sodium carbonate takes place in an iron
cylinder, placed horizontally and provided with agitators a,
Fig. 18. The charge is introduced at ^, and when finished,

Pig. 18

is drawn off to the filter at c. During the causticizing these
openings are closed with plugs. Steam may be blown in
through the pipe d to heat the liquor, while c is a rack for the
lime when this is used unslaked. In many works this rack
is dispensed with and the lime is slaked and screened from
lumps before going to the causticizer in the shape of milk of
lime. For this operation, sufficient sodium carbonate of
from 1.10 to 1.11 sp. gr. is run in so that when the lime is
added the causticizer will be nearly filled. At the ammonia-
soda works the liquor comes hot from the decomposition of

§ 39 ALKALIES AND HYDROCHLORIC ACID 77

the bicarbonate; in other cases it is better to heat it. Suffi-
cient lime is now added to fulfil the equation

Na^CO^ + Ca{OH\ = CaCO, + 'iNaOH

Theoretically, 106 grams of the sodium carbonate will
require 56 grams of calcium oxide, or G3. 3 grams of quick-
lime containing 90 per cent, of calcium oxide. Since the
above reaction is a reversible one, it is an advantage to have
an excess of lime present, so that about 10 per cent, in excess
of that theoretically required is employed.

The mixture is now kept, by blowing in steam, at a tem-
perature of about 80° C. and is constantly stirred by the
paddles for 3 or 3 hours, when about 92 per cent, of the
sodium carbonate will be causticized.

93. FUtratlon. — It is now necessary to separate the
caustic liquor from the calcium carbonate and other sus-
pended material (caustic mud), and although this is done at

some works by letting the liquor stand and settle, it is usually
filtered. For this purpose, such a filter as is shown in Fig. IH
is employed. It consists of an iron tank about 20 X 30 feet
and 4 or 5 feet deep. This tank is supported a little above
the floor by brick piers and the bottom is so sloped that
the liquor drains towards the pipe a. On the bottom of the
tank are strips b, cut out so that the liquor can circulate

78 ALKALIES AND HYDROCHLORIC ACID § 29

freely. On these strips are placed cross-strips c. These
cross-strips support bricks placed close together and then
comes a 6-inch layer of coke, about the size of hickory nuts,
followed by a 3-inch layer of finer coke, and then a thin
layer of clean sand. This is all covered with perforated iron
plates so that the workmen can shovel off the caustic mud
without disturbing the filter. The pipe a leads to the storage
tanks and, during the filtering, is under a vacuum. There
is a tendency for the caustic mud to crack and let the
liquid through unevenly, so during the filtering and washing,
workmen stir it occasionally with rakes. When the filtrate
has drained off, the caustic mud is well washed and the wash-
ings collected in a separate tank from the filtrate The
washings are used to dilute liquor for causticizing.

94. Evaporation. — The filtrate, which is mainly a dilute
solution of sodium hydrate, must now be evaporated, and in
the most economical manner, for the evaporation of such
dilute solutions is expensive at best. The proposition has
been made and carried out in some places to carry on part
of the evaporation in the steam boilers and then finally
run the stronger liquor to pots to finish the evaporation.

The evaporation of the caustic liquor in steam boilers has
several disadvantages and is for the most part abandoned. In
a few works the dilute caustic liquor is run at once into large
iron pots, which are heated by direct fire until the water is
all driven off. In the more progressive works, the caustic
liquors are brought up to about 1.3 sp. gr. by means of the
Yaryan evaporator and then run to the iron pots heated by
direct fire.

In the Yaryan evaporator the same principle is applied as
in the Pick evaporator for separating salt from brine. A
battery of Yaryans consists of three or four elements, which
are exactly alike, except that each following element works
under a lower pressure than the preceding one, so that
although the liquid in No. 2 element is more concentrated
than in No. 1, it boils at a lower temperature and therefore
can be boiled by steam from No. 1 element. In the same

§29 ALKALIES AND HYDROCHLORIC ACID 79

way, the steam from element No. 2 boils the caustic in ele-
ment No. 3, and so on. Usually only three elements are
worked together in a battery on account of the difficulty of
keeping the vacuum high enough in any more elements. So
far, the Yaryan apparatus resembles a large number of other
arrangements for working multiple effects. It is in the con-
struction of the elements, however, that the Yaryan' is
unique. Each element. Pig. SO, consists of an iron shell,

inside of which is arranged a number of sets of small cop-
per tubes — five or six tubes being in each set. The liquid
to be evaporated enters at e and is distributed to the sets of
tubes a, a. It circulates back and forth through these tubes
in the direction of the arrows, until it finally emerges against
the batHe plates in the space b b. Steam is meanwhile
admitted through f to the space between the tubes and
heats the contents to boiling. The admission of steam is so
regulated that all of it will be practically condensed to water

80 ALKALIES AND HYDROCHLORIC ACID § 29

in the apparatus and will finally flow away through g. This
condensed steam can be used for boiler feed water, or similar
purposes, if desired. The liquid flowing through the tubes
is in such a thin layer that as it boils it mixes with the
steam and fairly foams, so that the liquid comes in contact
with all parts of the tubes and gets the full benefit of the
heat, thus evaporating rapidly. The foaming mixture of
steam and solution issues from the tubes a, a^ and by stri-
king against the baffle plates in b b^ is separated. The solu-
tion settles and flows through c into the next element in the
series, where it goes through the tubes in the same way. The
steam passes upwards through the ** catch-all ** d^ where the
last of the particles of the solution, which are carried
mechanically by the steam, are separated and the solution
flows through // into the next element. The steam then
goes through /, which connects with f of the next element,
into the next element, and there boils the solution which it
has just left. This solution now passes through the tubes
of the next element under a lower pressure than it had in the
preceding case.

This system probably gives the most efficient evaporation
of any style of evaporating arrangement and is very compact,
as the elements can be placed one above the other. The
inventor of this apparatus claims that from 23 J^ to 25 pounds
of water can be evaporated with it in triple effect, and
30^^ pounds in quadruple effect per pound of coal, while in
the ordinary vacuum pan only 8^ pounds of wajer are
evaporated for the same amount of fuel. The apparatus
has the further advantage that it is nearly automatic in its
action, thus requiring but little attention, and since it con-
tains only a small amount of liquid at one time, it can be
easily stopped and started. The steam for the first ele-
ment is generated in a boiler kept for that purpose, but
for each following element it is supplied as pointed out
above.

95. Caiistlc Pots. — The evaporation cannot be success-
fully carried beyond a specific gravity of 1.3 in the Yaryan,

§ 29 ALKALIES AND HYDROCHLORIC ACID 81

for at this point the dissolved salts, such as sodium carbonate,
sodium sulphate, etc., begin to separate out. The solu-
tion is then run into the iron pots, where the evaporation is
finished. The salts that crystallize out are from time to
time ** fished" out, and the heating is continued until the
water is all expelled and fused caustic left in the pot. The
caustic pots are of cast iron and similar in shape to the
cast-iron pans used in making salt cake. They are ordi-
narily about 6 or 8 feet in diameter and from 3 to 5 feet
deep in the deepest part. The caustic pots are cast with
a rim, so that they can be supported on brickwork over a
grate, by which they are heated. A coal fire is generally
used for heating the pots, but since the fire must be
allowed to die down when the pot is finished, it has been
found very advantageous to use a gas fire for this purpose.
The caustic in the course of its evaporation attacks the
metal apparatus with which it comes in contact, so that by
the time it is finished the fused caustic contains copper, iron
oxide, and various other substances in suspension, as well
as aluminum, silicon, manganese, etc. in solution. The
substances in solution do not usually seriously affect the
caustic in value, although the manganese is frequently
plainly shown by the green manganate color. It is, however,
advisable to remove the suspended matter so far as possible,
and for this purpose, after all the water has been driven off,
the fires are cooled somewhat and the fused caustic allowed
to stand. The fused caustic is then ladled into sheet-iron
drums, which as soon as cold are sealed air-tight. In each
pot there is a residue, containing the settled impurities,
which is called the caustic bottom. The caustic bottoms
are put into drums and sold cheaply for making an inferior
grade of soap. When this is not possible, they are left in the
pots until they get too bad, when they are dissolved in water,
filtered, and reconcentrated.

96, Removal of Sulphur. — In the case of caustic made
from Le Blanc soda, the final removal of the sulphur
takes place in the pots. The sulphide is best oxidized to

82 ALKALIES AND HYDROCHLORIC ACID §29

thiosulphate, as already stated, by blowing in air. The
final oxidation of the thiosulphate is, however, very slow, so
that it is assisted by adding niter, a little at a time, until all
the sulphide and thiosulphate have been oxidized to sulphate.
Sometimes, instead of oxidizing the sulphide and so obtain-
ing it in a comparatively valueless form, the sulphur is
precipitated as zinc sulphide by using zinc oxide. The
reaction is

Na^S + H^O + ZnO = ^NaOH-^- ZnS

The zinc sulphide is separated before evaporating the
caustic liquor, and by calcination can be reconverted into the
oxide. After the removal of the sulphur, the caustic is
treated as in the above case.

97. Caustic Mud. — The material left on the filter in the
filtration of caustic soda goes by the name of caustic mud
and consists, especially when ammonia soda is used, princi-
pally of calcium carbonate. The composition of the caustic
mud from the filter of a works making caustic soda from
ammonia soda is given below :

CaCO^ 72.05^

Ca{OH\ 15.395^

Mg(PH\ 5.61j^

5/(9, 2.80j^

Fe,0, + Al,0 1.70^

CaSO, 29^

NaOH .48j<

H^O 1.62j^

99 . 94^

Many propositions have been made for utilizing this
material, among others, to use it instead of limestone in the
black-ash charge ; to use it for making Portland cement ; to
use it for whiting and to press it into form for crayon. It
has found some use in the still, instead of lime, to set
ammonia free from its salts ; probably, however, the greater
part of this material is still run to waste.

§ 29 ALKALIES AND HYDROCHLORIC ACID 83

98. Iioewlg'8 Process. — When sodium carbonate and

ferric oxide are mixed and fused together, carbon dioxide is

given off and sodium ferrite formed according to the

reaction

Na^CO^ + Fefi^ = %NaFeO^ + CO^

For calcination, a revolving furnace is usually employed
and the mass heated to a dull red. After fusion, the sodium
ferrite is allowed to cool and then washed with cold water
until all the soluble material is removed, then water of
80° to 90° C. is employed and the sodium ferrite decom-
posed into sodium hydrate and ferric oxide. The reaction is

The lixiviation can be so carried on that a caustic liquor
of 1.3 sp. gr. is obtained direct. This is the strength at
which the caustic leaves the Yaryan in the lime process, so
that a considerable saving is made in apparatus and fuel,
for this liquor can go direct to the pots. From that point
its treatment is the same as for caustic made from ammonia
soda by the lime process. The iron oxide used in this proc-
ess is a high-grade natural ore, as free as possible from
silica and other impurities, for these would lead to a loss of
soda through the formation of insolubfe compounds. The
iron oxide obtained by igniting precipitated ferric hydrate
is not suitable for this purpose, for on account of its fineness
it gives a product hard to lixiviate and filter. On the other
hand, the residue from the lixiviation of the sodium ferrite
can be used repeatedly and extra iron oxide is only needed
to make up for the mechanical loss. The process is not
especially valuable for making caustic from Le Blanc soda,
for the tank liquor must be evaporated and might as well be
causticized in solution and then evaporated. On the other
hand, it seems very well suited for working the solid
ammonia soda. Caustic soda of an excellent quality can be
made by this process.

99. Uses of Canstle. — Sodium hydrate is used prin-
cipally in the making of soap, wood pulp for paper making,

84 ALKALIES AND HYDROCHLORIC ACID § 29

and the purification of petroleum and other oils, although
considerable quantities are also employed in the purifying
of phenol and other organic substances. It is also now
used in considerable quantities in the preparation of coal-
tar dyes and in making sodium silicate and other sodium
compounds.

SODIUM BICARBONATE

100. The sodium bicarbonate of the ammonia-soda
process may be used just as it comes from the filters for
some purposes, but for most uses this is too impure. A few
years ago, practically all the sodium bicarbonate was made
direct from the Le Blanc soda crystals by spreading them
on racks and passing carbon dioxide over them. This had
the disadvantage of leaving all the impurities of the soda
ash in the bicarbonate, and later the method was improved
by dissolving the soda ash in water, or fusing it in its water
of crystallization, and passing in carbon dioxide. The bicar-
bonate then crystallized out, and most of the impurities
were left in solution. The soda from cryolite was especially
valuable for making bicarbonate on account of its great
purity. The making of sodium bicarbonate from ammonia
soda had the disadvantage for some time that it was difficult
to free it from ammonia. That difficulty has been over-
come, however, and at the present time practically all the
best bicarbonate of soda is made from the crude bicarbonate
of the ammonia-soda process. Two processes are in use
for purifying the crude bicarbonate, the zifit and the dry.

101, Wet Process. — This process consists in dissol-
ving the crude bicarbonate in hot water and saturating
the solution with carbon dioxide, then allowing it to cool
and the bicarbonate to crystallize out. The solution can
be heated to 05° C. without more than atmospheric pressure,
or to a higher temperature if a higher pressure is applied.
By this method almost all of the salt and other impurities
of the crude bicarbonate are left in solution. The recrys-
tallized bicarbonate is filtered off by means of a centrifugal

§ 29 ALKALIES AND HYDROCHLORIC ACID 85

machine and dried at a low temperature on traveling bands
of cloth.

102. I>ry Process. — This process consists in driving
off ammonia and moisture by a hot current of carbon
dioxide. It is not as good as the other, for it only removes
the volatile impurities from the bicarbonate.

However made, the bicarbonate is ground fine before
packing it for shipment.

The following analysis shows the high grade of purity
attained by the bicarbonate prepared from the ammonia-
soda crude bicarbonate. This sample was prepared by the
wet method, which is the one most used.

HNaCO^ 99.400^ Alfi^ + Fefi^ . .009j<

Na^CO^ 880^ Na^SO, 007j^

NaCl. 023^ CaCO^ 021^

SiO^ 008^ MgCO^ 011^

ALKALIES AND
HYDROCHLORIC ACID

(PART 2)

CHEMICAL METHODS

HYDROCHIiORIC ACID

PROCESS OF MANUFACTURE

1. The manufacture of hydrochloric acid is almost insep-
arably connected with the manufacture of salt cake, and
really consists in the condensation of the acid set free in the
salt-cake manufacture. In a few works salt is decomposed
for the hydrochloric acid alone ; in which places the charge
of salt is always in excess of the sulphuric acid, for the salt
is much cheaper than the acid and the more expensive sul-
phuric acid is more completely utilized than it is in the
ordinary salt-cake process. A purer hydrochloric acid is
also obtained in this case. The apparatus and methods of
working are, with the above exception, the same as in the
making of salt cake. We shall, therefore, merely consider
the condensation of the hydrochloric acid that has been
made in the salt-cake manufacture.

2. CJondensatlon of Hydrochloric Acid. — During the
early years of the manufacture of salt cake, the hydrochloric

§30

For notice of copyrightt see page immediately following the titie page.

2 ALKALIES AND HYDROCHLORIC ACID § 30

acid had very little value and was allowed to escape freely
into the air. But the action o£ the gas was so bad on vege-
tation and, although it has not been proved that it has an
injurious effect on animals and men, it became such a nui-
sance as the works increased in number that, in 1862, the
Lord Derby Alkali Act was passed in England forbidding
manufacturers to allow more than 6 per cent, of their hydro-
chloric acid to escape into the atmosphere. The present
English Alkali Act only allows .2 grain of hydrochloric acid
per cubic foot of chimney gas to escape into.the atmosphere.
This makes it necessary to absorb the acid in water; we
have passed also from the time when the salt cake was a
source of profit and the acid a troublesome by-product to
the time when the acid is the chief source of profit. The
problem that now confronts the manufacturer is how to get
the most complete absorption of the hydrochloric acid in the
cheapest manner, and at the same time make the strongest
solution of the acid possible. Hydrochloric acid being a gas,
its concentration in solution depends on the temperature and
pressure; under ordinary conditions, the strongest acid is
about 40 per cent. , while in practice the best working gives
about 36 per cent, in winter and 30 or 32 per cent, in summer.

3. Usually the pan acid is absorbed separately from the
roaster acid, for the pan gases contain a comparatively high
percentage of hydrochloric-acid gas and can be more easily
absorbed to a strong acid solution, while the roaster gases
tend to give a much weaker and more impure acid. The
gas from the pan is cool enough, so that it can be conducted
away in glass or earthenware pipes; the gas from the
roaster, however, is so hot that it would crack these pipes
and either brick flues or iron pipes are used. The brick
flues are disadvantageous, because they do not permit rapid
cooling; iron pipes are much better, for they permit very
rapid cooling, but they cannot be used after the tempera-
ture of the gas gets below 200° C. After the roaster acid
is fairly cool, it receives the same treatment as the pan acid,
so that they will be considered together.

^ - ■ ■ ■ ' J''.V- ■ " ■. ...J. >•: — . .. .1 ■ ■ '.

• :■. •.•■■■■ ...' .■-■.■*..■.■ ».••'*." • ■■■
;>'••.',•' • '.■■.'•■..: • '.'.:'■;■■"[' '•■'.'.'.{:

I"'' ■' ■ ■-• '. .'•■ "j^'t" •■ . "'

. ■ 1 .^ . . i ;

' v ■ ■

\--V::;:.^.v;-V;;.;.;^;;^-.^V.V'V----v^-V--V"^\V'-'=:'" —

■•.•..•.^..^.•\!..-...^X%..;;;-..V:.,.-.\..-\v:..-V.:-— -^V

■ .•..•'^.••Jv.\v..- ■• ■ li--

§ 30 ALKALIES AND HYDROCHLORIC ACID 3

Cold hydrochloric acid absorbed in cold water will gener-
ate enough heat at a 16-per-cent. solution to boil water, and
at a 20-per-cent. solution to boil hydrochloric acid of that
strength ; so, if there is no outside cooling, 20 per cent, is
the highest strength of the acid possible. It is, therefore,
necessary to furnish the system with an efficient cooling
arrangement, although it is now generally recognized that
it is better to first saturate the gas with water vapor and
then condense the mixture to as strong a solution as possible.
The essential points to be borne in mind in condensing
hydrochloric acid are, therefore, to cool the gas thoroughly,
keep it cool throughout its condensation, and to bring it
into intimate contact with the absorbing water.

4. Apparatus. — The kinds of apparatus used for the
condensation of hydrochloric acid and the arrangement of
( the same have gone through several stages of development,

! until today the practice is quite varied in this respect.

I The following arrangement shows most of the various types

of apparatus in their best forms and it gives the most satis-

I factory condensing arrangement at present used.

J Fig. 1 shows an elevation and ground plan of this system.

The gas goes from the pan and roaster through the pipes A

I and A'. The pipe from the pan is made of earthenware

tubes tapered so that the small end of one fits into the large

I end of the next, and so on. The pipe from the roaster is

made of iron for one-half its length and of earthenware for
the remainder. The conducting pipes are not made very long,
as their function is to conduct the gas to the towers B and
B' and not to cool it, although the gas is somewhat cooled
in passing through them. The conducting pipes are sloped
downwards to the bottom of B and B\ so that any acid
condensing in them will run to the bottom of these towers.
The towers are made of stoneware and are about 4 feet
square and 12 feet high ; the lower half is empty and the
upper half filled with fireclay cylinders set on end. Water
is allowed to flow down these towers in such amounts that it
is practically all vaporized by the hot acid gas. This water

ALKALIES AND HYDROCHLORIC ACID § 30

serves the double purpose of washing the sulphuric acid
from the gas and of cooling it. The fairly cool gas, satu-
rated with water vapor, now enters the bombonnes C and
C\ where it meets a stream of water flowing in the opposite
direction to the flow of the gas. This water and the tall
connecting pipes of the bombonnes finally condense most of
the hydrochloric-acid gas; a certain amount of the acid
always escapes condensation here, however, and is removed
by the coke tower D,

The bombonnes, Fig. 2, are made of earthenware and
are fitted with rather long earthenware pipes a to cool the
N gas, as well as to conduct it

from one bombonne to the
next. The bombonnes are
also connected at b by ground
joints, or a glass tube and
rubber stoppers ; on the inside
of each bombonne there is a
pipe c to conduct the incoming
liquid to its bottom and natu-
rally the upper portion flows
on to the next bombonne in
order. From twenty to thirty
of these bombonnes are used in series for each pan.

The coke tower is made of acid-proof stone slabs fastened
together by iron bands soaked in tar ; it should be about
5 feet square and 30 or 40 feet high. As the name indi-
cates, the packing used is usually coke, but porous stone and
various forms of earthenware have been used. The towers
are best used in pairs, the acid gas entering at the bottom
of one, and rising to its top, is carried by a pipe to the
bottom of the next tower, and escapes at the top of this.
Water constantly flowing down over the coke absorbs the
acid. The packing of the tower requires considerable
attention; for if the packing is too loose or the pieces of
coke are too large, not enough surface is offered for the acid ;
on the other hand, if the packing is too tight there is not
enough draft. The coke used in packing the towers should

Fig. 2

g 30 ALKALIES AND HYDROCHLORIC ACID 5

be the hardest oven coke. In the bottom of the tower the
largest and longest pieces should be used and smaller pieces
in order, until, after an eighth of the way up, pieces 6 or
8 inches by i inches mixed with some smaller ones can be
used. After one-third of the tower is carefully packed, the
rest can be filled by dumping in coke, freed from all pieces
under % inches by riddling.

5, Ijunsre Plate To'wer.— Another form of coke tower,
or condenser, is obtained by using the Lunge plates. The
plate tower. Fig. 3, only occupies from
^ to ^ the space required for coke
towers, and gives an even more efficient
absorption for the gas. Of course, the
size of the tower will vary with the
work required of it, but for ordinary
cases the best tower consists of nine
earthenware cylinders, each 3 feet in
diameter and 3 feet 3 inches high, set
together as indicated in the figure.
The first cylinder A is left empty, the
next three i are filled with sixty Lunge
plates, the next one c is left empty, the
next two t/are filled with coke, and the
last two e are empty. Whatever size
of tower is used, this is the best distri-
bution of the filling. The gas passes in
at g; meets a descending stream of
water, which absorbs the hydrochloric
acid, and flows out at A into the bom-
bonnes. The waste gases pass out
through /.

6> Hart System. — A system re-
cently patented by Hart for the
absorption of hydrochloric acid has "*'

much to recommend it in compactness and simplicity, and
although it has not been used hmg enough to warrant
its being called an establisihed method, it deserves some

6 ALKALIES AND HYDROCHLORIC ACID § 30

consideration. It consists of a series of glass pipes a, Fig. 4,
through which water runs. The water Is fed in continu-
ously at d and flows from one pipe to the next until it finally
runs into c as strong acid. These pipes are cooled by water

running over them from a perforated pipe </, the excess
being carried off in e. The gas comes in at /, passes over
the strong solution of acid in c, and then through the pipes
to the flue A, where it goes to the chimney.

7. Commercial Hydrochloric Add. — This is a yellow-
colored solution of the gas in water, usually claiming a
specific gravity of 1.3, but rafely containing over 30 or
35 per cent,, by weight, of hydrochloric-aci^ gas, and
seldom, if ever, reaching so high a concentration as
40 per cent, of the acid. Its yellow color is mainly due to
organic matter, for it seldom contains enough iron to
seriously atfect its color. It contains, as other impurities,
sulphuric acid, chlorine, arsenic, and frequently lead and
calcium chlorides.

§ 30 ALKALIES AND HYDROCHLORIC ACID 7

8, Purification of HydrocWorlc Add. — For many
purposes, the crude hydrochloric acid will answer very well,
but for others it must be as nearly chemically pure as possible.
For this reason a method of purification that is suitable in
one case will be useless in another ; and, furthermore, the
question of cost must frequently be taken into consideration.
An adoption of one of the following methods will usually
meet every demand.

The cheapest and most effective method for purifying hydro-
chloric acid, especially from sulphuric acid, is the so-called
Hasenclever method. This consists in treating the strong
water solution with concentrated sulphuric acid or calcium
chloride and blowing air through the mixture. The hydro-
chloric acid is evolved free from practically all the impuri-
ties except the arsenic, and may be used in the gas forni, as
is usually done, or reabsorbed in water. In carrying out
the Hasenclever method, 100 parts of the crude hydrochlo-
ric acid is run into a stone jar with 550 parts of sulphuric
acid of 60.4° Baum6, and the mixture stirred mechanically or
by means of a current of air when there is no objection to
having air mixed with the hydrochloric acid. The sulphuric
acid is thus reduced to 55** Baum6 and is reconcentrated by
surface heat.

Sulphuric acid and sulphur dioxide may also be cheaply
removed from hydrochloric-acid gas by passing it through
towers containing solid sodium chloride. Where arsenic-
free acid is needed, it is best to start with arsenic-free sul-
phuric acid. If the acid is diluted to 1.12 sp. gr. and
barium sulphide is added, the arsenic will be precipitated as
the sulphide and the sulphuric acid as barium sulphate.
The gas may then be distilled off and reabsorbed in
water.

Another method consists in adding a solution of stannous
chloride in concentrated hydrochloric acid to the strong
hydrochloric acid. Arsenic separates out, the reaction
probably being

lAsCl^ + ZSnCl^ = 2^j + ZSnCl^

8 ALKALIES AND HYDROCHLORIC ACID § 30

This leaves stannic chloride in the acid unless it is redis-
tilled.

Arsenic and chlorine may be removed by digesting the
acid with scrap copper for some hours; the arsenic is pre-
cipitated and the chlorine combines with the copper. The
acid is then redistilled.

9. Uses of Hydrocliloric Acid. — About three-fourths
of all the hydrochloric acid made is used in the preparation
of chlorine. The remainder is used for making the chlorides
of various metals, various acids, as carbonic, etc., gelatine,
superphosphates, for purifying animal charcoal, in dyeing
and bleaching, in the manufacture of dyestuffs, for the
preparation of various food products, in various metallur-
gical operations, etc.

CHLORINE

lO. Historical. — About the time that soda ash was
beginning to be made by the Le Blanc process, Scheele (1774)
found that by certain reactions he could obtain a new sub-
stance from hydrochloric acid. He did not consider that
this new gas was an element, but called it **dephlogisticated
muriatic acid." Even after the phlogiston theory had been
disproved, the idea still prevailed that an acid must contain
oxygen, and that since this new gas was made by taking
hydrogen away from muriatic acid, it must also contain
oxygen. It was not until 1810 that Davy succeeded in
proving the elementary character of chlorine, and this view
was not accepted by Berzelius until 1821.

In 1785 Berthollet recognized the bleaching effect of chlo-
rine on cloth and proposed its use on a commercial scale.
He advised using chlorine water for this purpose. The
chlorine water did not keep well, however, and its prepara-
tion on a large scale was not convenient ; so in 1789 the plan of
passing the chlorine into a solution of potash was originated
at the Javel works, near Paris. In this manner, potassium
hypochlorite, known as Fmu dc Javel ^ was made.

§ 30 ALKALIES AND HYDROCHLORIC ACID 9

Early in 1798 Charles Tennant, an Englishman, tried to
patent a process for absorbing chlorine in milk of lime, but
the patent was not allowed on account of having been antic-
ipated by some one. In April of the next year, however, he
patented the absorption of chlorine by dry, slaked lime, and
so established the making of bleaching powder by our present
method. During 1799 he made 5*^ tons of bleach, which he
sold at $700 a ton'; this is striking contrast to the large
amoimt now turned out every year, and selling at an average
of $25 or $30 a ton, or even less.

11. Sonrce. — Just as sodium chloride is the substance
from which practically all the sodium carbonate of <:om-
merce is^ made, so it is also the chief source of chlorine.
Potassium chloride and magnesium chloride furnish a small
supply, and calcium chloride and some other chlorine com-
pounds have been proposed as suitable material for the fur-
nishing of chlorine, but the problem of getting the chlorine
from these substances in a commercial way has not yet
been solved.

13, Clilorlne Direct From 8alt. — In spite of the fact
that the making of chlorine and sodium carbonate began to
be important commercially at about the same time, and
that the manufacture was frequently carried on by the same
firm, and usually in the same locality, the chlorine was made
direct from salt, and the hydrochloric acid from the salt-
cake furnaces allowed to go to waste and become a nuisance
in the neighborhood. The operation of making chlorine
consisted in mixing salt and manganese dioxide and treat-
ing the whole with sulphuric acid. This brought about the
reaction

^NaCl+ MnO, + d//^SO^ = MnSO^ + 2Na/fS0, + 21/^0

and all the chlorine was obtained from the salt, but at the
expense of large quantities of sulphuric acid. A portion of
this sulphuric acid can be saved if the temperature is kept

10 ALKALIES AND HYDROCHLORIC ACID § 30

high enough to drive the reaction to the formation of the
normal sodium sulphate. The reaction then becomes

^NaCl + MnO^ + 2//,50, = Na^SO^ + MnSO^ + ^Hfi

+ Cl,

This reaction is only obtained, however, at a temperature
above 120° C. , which is not easy to obtain with steam, and
any other method of heating is almost out of the question
on account of the material necessarily used for decomposi-
tion vessels. This process is still sometimes carried out in
the chemical laboratory and at a few places where chlorine
is only needed in comparatively small quantities.

13« CMorine From Hydroclilorlc Add. — ^When the
Le Blanc soda works began to increase in size and number,
the escape of the hydrochloric acid into the air became such
an unbearable nuisance that it had to be abated by absorb-
ing the acid in water. This soon made hydrochloric acid
abundant and cheap, so that it then came into use for
making chlorine. The preparation of chlorine from hydro-
chloric acid consists essentially in the removal of the hydro-
gen from the acid by an oxidizing agent. In selecting the
oxidizing substance, its cheapness and efficiency must both
be taken into account, as well as the ease in handling and
the resulting products. Naturally, an oxidizing substance
that can be easily and cheaply regenerated by means of the
air is much to be preferred to one that must be thrown
away when once used.

14. Oxidation by Oxides of Mangranese. — Just as
oxides of manganese were used to act with salt and sulphuric
acid for the preparation of chlorine, so they have been used
more recently with hydrochloric acid for the same purpose,
for they occur in nature in large quantities, but in varying
states of oxidation. The oxides of manganese occurring in
nature are manganosite MnO and pyrolusite MnO^y which
represent the high and low degrees of oxidation, and the
intermediate oxides braunite Mn^O^y manganite j^«,d>„//,(?,

§ 30 ALKALIES AND HYDROCHLORIC ACID 11

hausmannite Mnfi^^ wad, and psilomelane. The last two
contain the manganese mostly in the form of manganese
dioxide, but also contain varying quantities of other metals.
The reactions that occur between the oxides of manganese
and hydrochloric acid are as follows :

MnO + %HCl = MnCl^ + Hfi
MnO^ + 4//a = MnCl, + 2//,(? + C/,
Mn^O, + ^HCl = %MnCl^ + 3//,(? + C/,
Mnfi^ + %HCl = ^MnCl^ + A^H^O + Cl^

It will be readily seen that manganese dioxide yields the
highest amount of chlorine for a given amount of hydro-
chloric acid, and that the presence of other oxides, as well
as of iron, calcium, and other metals, is a disadvantage, as
it lowers the oxidizing power of the ore and uses acid to no
purpose. The manganese ore is usually bought according
to its percentage of available oxygen, which is con-
sidered to represent the amount of manganese dioxide in
the ore.

The hydrochloric acid is, of course, used in solution, and
the stronger the solution, the better it is. At best, only
50 per cent, of the acid in the solution can be made to
yield chlorine, as will be seen from the second reaction.
The reaction does not continue after the strength of the
acid has fallen to 5 per cent, and usually, under ordinary
working conditions, 7 or 8 per cent, of the acid is left in the
residual liquors. These latter percentages do not mean those
of the acid originally present, but are the actual percentages
of acid in the solution, so that it is easily seen that a far
greater percentage of the acid is left unused when an acid
of 10 per cent, original strength is used than when one of
35 per cent, is employed. Working under the best condi-
tions by this method, rarely over 30 or 33 per cent, of the
total chlorine of the acid is obtained in an available form.

16. Apparatus. — The stills for the decomposition of
hydrochloric acid by means of manganese dioxide are made

either of earthenware or of
been boiled in tar to make it

13 ALKALIES AND HYDROCHLORIC ACID §30

silicious sandstone, which has
acid proof. A small still used
in works of limited capacity,
and sometimes also in larger
establishments, is shown in
Pig. 5. It is made from
sandstone and consists of
two parts joined by a
tongue and groove with
rubber cement. A little
above the bottom of the
still is a narrow ledge upon
which rests the perforated
section b. The manganese
ore in small lumps is
placed on b and the hy-
drochloric acid run in
through d. The chlorine
gas, as it is evolved, passes
out through e, and as the
n into the still through c, and

action slackens, steam is i
coming out under the
false bottom, mixes and
heats the contents of
the still. At y is a man-
hole, which serves for
introducing the man-
ganese ore and for
cleaning the still; the
residual liquor is drawn
off through i"-.

In Fig. G is shown
another form of still,
which is very suitable
for the preparation of '
chlorine on a small
scale, although it is
hardly suited for larger

§ 30 ALKALIES AND HYDROCHLORIC ACID 13

works. It consists of a sandstone or earthenware still a,
provided with a false bottom A, as in the above case. The
still a is set in a wooden case and surrounded by a concen-
trated salt solution, which serves as a lute for the bell r.
This bell is suspended by chains on pulleys and counter[K>ised
by weights, so that it can be easily moved up and down as
desired. In the top it is provided with a funnel tube (/for
the introduction of the acid, and an exit tube f for the
chlorine. The spent liquor is drawn off through / By
blowing in steam through ^, the salt solution can be warmed
as desired and the contents of the still brought to the
desired temperature without diluting the still liquor by
blowing in steam.

Another form of still, which is shown in Fig. 7, is made
from sandstone slabs grooved together and made tight by

means of rubber cords that fill the connecting grooves. These
stills work on the same principle as those just described, but
are much larger and better suited for work on a large scale.
In this apparatus the lumps of manganese ore rest on the
false bottom a and the acid is run in through <*. The lower
end of the tube e: dips into a cup, which is continually
filled with hydrochloric acid, and so forms a lute to prevent

14 ALKALIES AND HYDROCHLORIC ACID § 30

the chlorine escaping through the tube. Steam is introduced
through b when necessary, and the chlorine escapes through d.
The waste liquor can be drawn off through e into a trough
and run away.

16, Managrement of tlie Stills. — The operation con-
sists in charging with manganese ore and then running in
hydrochloric acid as rapidly as the reaction will permit. The
evolution of chlorine is allowed to continue without heat for
from 8 to 12 hours, when steam is blown in at intervals.
Steam cannot be blown in continuously, for the temperature
would become too high and too much hydrochloric acid and
water would be carried over into the chlorine. The chlorine
would also be likely to come off too rapidly. The pipes for
conducting the chlorine are either of lead or earthenware and
the gas is often conducted from several stills into one large
main pipe. In this case, when a still is stopped to be cleaned
and refilled it is necessary to cut it off from the main pipe.
This cannot be accomplished by using valves or stop-cocks,
but is brought about by a variety of means, two of which are
described here.

One of these methods is shown at d^ Fig. 7. The con-
ducting tube is connected with a Y tube shortly after leaving
the still, and this Y tube, which is open at its lower end, sets
in a jar, as shown. When it is desired to bring the tube into
action, the liquid in the jar is lowered to below the branching
point of the Y, but its lower end is left covered. The chlorine
cannot escape to the outside, but can easily pass through the
branches of the Y to the large conducting main. When it is
necessary to close the tube, it is easily done by filling the jar
with water or a solution of salt. The branches of the Y will
then be filled and the passage stopped. With this arrange-
ment, it is necessary to empty or fill the jar each time a
change is desired, which is inconvenient.

A much better arrangement consists in making a U bend
in the tube. Fig. 8. At the lower end of the U a small tube
is connected, to which may be fastened the flexible tube a.
The tube a is connected to the cup b^ which contains a strong

§30 ALKALIES AND HYDROCHLORIC ACID 15

salt solution. When b is raised, the solution flows into the
U tube and shuts off the flow of gas; when b is lowered, how-
ever, the solution flows out of the U tube into b and the
: is open for the gas.

17. StlU Uqnors. — The liquors from the stills contain,
in the form of chlorides, all the manganese, aluminum, iron,
calcium, etc. that were contained in the ore, together with
considerable hydrochloric acid. Although the liquor varies
considerably with the grade of manganese ore used and the
strength of the hydrochloric acid, the following may be
considered as a fairly representative analysis:

HCl 6.62-^ FeCl, 46^

AlCt^ &%^ Hfi 81.73^

MnCl^ 10.57^

This liquor, on account of the large amount of acid that it
contains, is hard to dispose of, for if given a chance it will
act on the mortar in the foundations of buildings and even
on the stones themselves. If run into the streams, it kills
the fish and acts in a generally disagreeable manner. How-
ever disposed of, when it is flrst run from the still it evolves

16 ALKALIES AND HYDROCHLORIC ACID § 30

a disagreeable odor of chlorine. In addition to all these bad
qualities, the still liquor also carries away with it all the man-
ganese, and as manganese ore began to be scarce and the
price to increase, a method for treating these liquors became
almost a necessity. Of the large number of processes pro-
posed for this purpose only one will be described here.

18, Weldon's Process. — It has long been known that
manganese hydrate can be precipitated by lime water and
that it is somewhat oxidized by the oxygen of the air. All
attempts to utilize these facts for the recovery of manganese
were, however, for a long time futile, for the oxidation
proceeded too slowly and could only be driven to the for-
mation of Mnfi^^ or at best Mn^O^, It was only when
Weldon discovered that with an excess of calcium hydrate
the oxidation went on more rapidly and to a greater
degree that the process had any commercial possibilities. It
is now practically the only process used for the recovery of
manganese, and it figures in the preparation of a large per-
centage of all the chlorine made. The process consists in
first neutralizing the still liquor with powdered chalk. An
excess of chalk is to be avoided, as in settling it increases the
precipitate and so increases the loss of manganese. Sufficient
chalk has been added when the liquor no longer gives an acid
reaction with litmus paper. The neutralized liquor is then
run into settling tanks, where the excess of chalk and the
iron and aluminum hydrates are allowed to deposit. The
clear liquor is then run into the blowers, where it is
heated to 55° C. and mixed with enough calcium hydrate
to precipitate all the manganese as hydrate and then from
one-fifth to one-half more of the lime is added. The
calcium hydrate used for this purpose should be as pure as
possible; it must especially be free from magnesium com-
pounds, for the magnesium chloride is not decomposed by
the chalk in the neutralizing tank, but goes to the oxidizer,
where it is precipitated by the lime and goes on to use up
hydrochloric acid at a later stage of the process. As soon
as the manganese hydrate has been precipitated and a

§ 30 ALKALIES AND HYDROCHLORIC ACID 17

proper amount of lime in excess is present, air is forced
through the mixture and the oxidation begins. The air is
blown through the apparatus for from 2^ to 4 hours, depend-
ing on the apparatus, although at the same works the time
for blowing is about the same for each batch. At the end of
the first blow, a calcium manganite of practically the com-
position CaO{Mn0^^y together with other manganites, is
formed. Then, without stopping the blowing, a suitable
amount of manganese chloride (about one-fourth the amount
originally taken) is run in and the blowing continued until
this is oxidized as far as possible.

The oxidizing of the manganese hydrate requires con-
siderable care and experience, for the blower must be started
at exactly the right time and at the proper speed. If it is
started too strongly before a sufficient excess of lime is
added, the manganese is oxidized to Mn^O^, and after this
is once formed it is very difficult to force the oxidation any
farther. Such a result is called a '*red " or ** foxy " batch
on account of its being a brownish-red color instead of black,
as it should be. On the other hand, if the blower is not
started quickly and strongly enough, the contents of the
oxidizer become thick, so that it is very difficult to force
the air up through it; such a result is called a ** stiff batch."
The only remedy is to start the blower at full strength and
carry the batch by this point, if possible. A stiff batch may
also be caused by too high a temperature or too little calcium
chloride in the mixture. The best mixture contains about
3 gram molecules of calcium chloride to each gram mole-
cule of manganese chloride. For the total oxidation, it is
estimated that 300,000 cubic feet of air are required to
recover the manganese for each ton of bleach made. At
many works the addition of the manganese chloride and the
continuation of the oxidation is not practiced. That it is
advisable, however, is shown by a consideration of the reac-
tions taking place in the Weldon process.

19. Reactions. — If we leave out of consideration the
neutralization of the still liquor, which really is not one of

18 ALKALIES AND HYDROCHLORIC ACID § 30

the parts of the process proper, the first reaction is the pre-
cipitation of the manganese hydrate, which, if represented
for 100 gram molecules of manganese chloride, is

lOOMnC/^ + 100Ca{O//)^ = 100Mn{ON)^

+ lOOCaC/^

For oxidation, the extra lime is added, as mentioned
above, and air blown in; the reaction then taking place,
neglecting the nitrogen of the air, may be represented thus :

100Mn(O//)^ + 60Ca(O/f)^ + 43(9, = 4%CaO,MnO^
+ l^MnO.MnO^ + nCaO{MnO;)^ + IQOH^O

This is, if we consider the oxidizing and basic parts sepa-
rately, equal to S6MnO^ + 74:{CaO + MnO). Thus, out of
100 gram molecules of manganese chloride we get 86 gram
molecules of MnO^^ or active material for oxidizing the
hydrochloric acid; but we have also 74 gram molecules of
substances that neutralize and so destroy hydrochloric acid
and yield no chlorine. Now, if we add an extra quantity of
manganese chloride, a part of the above material reacts, and
we get

+ 2^Mn{0H)^ + UCaCl^

Then, by blowing, the manganese hydrate is oxidized
according to the equation

24:Mn{OH)^ + 6(9, = lUfnO.MnO^ + UH^O

By collecting the above equations and adding them alge-
braically, we have

XWiMnCU + 'iOOCa(0//), = \^Mn(OH\ + lOOCaC/t
100Mn{O//)9-\-e0Ca(ON)^ + 4»Ot = 4SCaO,AfnOt + UMnO.MnO^

+ \2CaO(MnOt)t + l^H^O
UH^O-¥^CaO,MnO^-¥UMnCU = UCaO,{MnOt)^ + UMn(OH)^

H- 24CaC/,
^AMn{pH)^ + 60, = nMnO.MnO^ + 24^,0
ViAMnCl^ + \mCa{OH)^ + 49(9, = ^CaOAMnO^)^ + ^MnO.MnO^

H- ViACaCU + X^H^O

From this last equation it may be noted at once that at
the end of our operations we have a mixture of Z^CaO^

§ 30 ALKALIES AND HYDROCHLORIC ACID 19

{MnO;)^^nd26MnO,MnO^, or 9SMnO^+ S6CaO+26MnO,
from 124 gram molecules of manganese chloride. That is,
we have 79 per cent, of the manganese in the form of the
dioxide, as against 86 per cent, of the manganese in this
condition before the last addition of manganese chlo-
ride. The present condition is much better, however, for
although the percentage of the manganese converted to the
dioxide is somewhat smaller than before, the amount of base
present is much more reduced than the active manganese.
Before the second addition of manganese chloride, there are
74 gram molecules of base to 86 gram molecules of manga-
nese dioxide, that is, 53.75 per cent, of the total number of
gram molecules that can react with hydrochloric acid is
madganese dioxide. When the operation is completed, how-
ever, there are only 62 gram molecules of base to 98 gram
molecules of the manganese dioxide, or 61.25 per cent, of
the active gram molecules is manganese dioxide. It is
obvious, then, that the second addition of manganese chlo-
ride and longer blowing is a decided advantage.

30, Weldon Mud. — The mixture of calcium and man-
ganese manganites obtained by the above operations is a
black, shiny precipitate, which is in suspension in a solution
of calcium chloride. This mixture is run from the oxidi-
zers to the settling tanks, where it is allowed to stand for 3 or
4 hours. At the end of this time the precipitate will have
settled into the lower half of the solution and the clear cal-
cium chloride solution can be drawn off from the top; the
shiny mass remaining is called Weldon mud. The Weldon
mud finds several uses besides the preparation of chlorine ;
it is used in gas purifiers, to remove iron from alum, to
remove sulphides from caustic soda, and for several similar
purposes. Weldon at one time recommended it for neu-
tralizing the still liquors instead of chalk, but later aban-
doned it for that purpose. At present it is used quite
extensively in that way, for it not only saves the chalk, but
also utilizes the acid of the liquor to neutralize the bases in
the mud, and so increases the efficiency of the mud as an

20 ALKALIES AND HYDROCHLORIC ACID § 30

oxidizing agent. The use of Weldon mud for neutralizing
the still liquors has the disadvantage that all the impurities,
such as calcium sulphate, iron, and aluminum, are left in
the mud. This makes it necessary to occasionally neutral-
ize a batch with chalk and allow the impurities to settle
out. When this method is used, great care is taken to keep
sulphuric acid out of the hydrochloric acid. Sometimes
calcium chloride is added to precipitate the sulphuric acid
before the hydrochloric acid is used.

CHLORINE BY THB WELDON PROCESS

31, Of course the chief use for Weldon mud is the gen-
eration of chlorine, and for this purpose it is much more
active than manganese ore. The stills used are similar to
those already described, but the method of working is some-
what different from that when manganese ore is used.
In working with the Weldon mud, the hydrochloric acid
is run as hot as possible directly from the condensers into
the stills, and the Weldon mud is then added slowly, so as
to regulate the flow of chlorine until sufficient for the acid
is present. Too much must not be added, especially if the
still liquors are neutralized by chalk, for in that case the
manganites, that are unacted upon, will settle with the mud
from the neutralized liquors and be lost. When the color
of the liquor in the still shows that enough mud has been
added, steam is blown in and the chlorine driven off as com-
pletely as possible. It is possible in this way to leave
only from ^ to 1 per cent, of free hydrochloric acid in the
still liquor. This is equivalent to about 3 per cent., as
counted on still liquor from manganese ore, for the water
in the Weldon mud makes its still liquor more dilute than
that from manganese ore. From 1^ to 3 per cent, of the
manganese is lost in the cycle of operations, and this is sup-
plied by continuously decomposing the necessary amount of
manganese ore in a small still and adding its liquor to the
general supply. Only about 30 per cent, of the chlorine in
the hydrochloric acid is obtained in the bleaching powder.

§30 ALKALIES AND HYDROCHLORIC ACID 21

The remainder is, for the most part, run to waste as calcium
chloride.

33. Apparatus. — The apparatus by means of which this
cycle of operations is performed is shown in Fig. 9, which
represents a cross-section through part of it. Starting with

the still liquor from the still A, the liquor runs into the neu-
tralizing tank B, where it is mijtedwith chalk or Wcldon
mud and thoroughly stirred. It is then pumped, by means
of the pump C, through the pipe shown, to the settling
tank D. If chalk has been used for neutralizing, the mud
obtained is valueless. If Weldon mud was used, however,
the mud here obtained can be used in the chlorine still.

22 ALKALIES AND HYDROCHLORIC ACID § 30

From D the neutralized liquor goes to the oxidizers Ey E,
Lime is, meanwhile, slaked in F, F, and made to the proper
consistency. It is then pumped, by the pump and pipe
shown, to the reservoir (7, from which place it is run in
proper quantities into the oxidizers E^ E, Air is forced
into the mixture through the pipe /, which extends to
the bottom of the oxidizers, by the blowers H, From the
oxidizers the batch is drawn off into the settling tanks Ky
from which the mud is again run as needed into the still A,
It will be noted that almost all the materials are moved as
solutions or slimes, so that the work is almost entirely
mechanical. The solutions or slimes are pumped to the
highest point of the plant and then allowed to flow down
through the various pieces of apparatus until they once
more reach the lowest point. Practically the same number
of men are required for a small plant as for a large one, so
that the working of a large plant is on thi^ account more
economical.

BEACON'S PROCESS FOR CHLiORINE

33, In the process just described the manganese has
acted simply as an oxidizing agent to remove the hydrogen
from the chlorine and set the latter free. Although the
steps are a little farther removed, there is a direct analogy
between this operation when the Weldon manganese-recov-
ery method is employed and the making of sulphuric acid
where nitric oxide is used as*a carrier of oxygen from the
air. And, just as recently the problem of causing sulphur
dioxide to combine directly with the oxygen of the air by
passing a mixture of the two gases over platinized asbestos
or ferric oxide has been solved in a practical manner, so,
much earlier, it was found that when hydrochloric acid and
air are passed over porous material saturated with salts of
copper, lead, or manganese the oxidation of the hydrochloric
acid takes place direct.

It was discovered and patented by Oxland in 1845 that
when a mixture of hydrochloric acid and air is passed

§ 30 ALKALIES AND HYDROCHLORIC ACID 23

through a tube filled with red-hot pumice, the following
reaction takes place :

4Ha+ O^ = %H^O + 2C/,

This is a reversible reaction, however, and, under the
conditions here stated, the decomposition of the hydrochloric
acid is very incomplete. Ten years later, 1855, Vogel found
that when cupric chloride is heated it decomposes into
cuprous chloride and chlorine according to the reaction

2CttC/, = 2C«C/+ C/,

Then by passing hydrochloric acid and air over the
cuprous chloride, an oxychloride of the composition
CuCl^^ZCuOyZHfi is formed, which finally goes over into
cupric chloride, the final reaction being

It was found, however, that in practical working only
about one-third of the chlorine was obtained from the cupric
chloride, instead of the theoretical one-half. There was
also a loss of copper salts, and on account of these and other
difficulties, the process was never successful.

The idea occurred to Deacon, however, to combine these
two methods, and he took out his first patent to that effect in
1868. Variojus contact substances have been proposed and
patented, but certain salts of copper are found to be the best.
In general, the process as carried out now consists in passing
a suitable mixture of hydrochloric acid and air through tubes
containing clay balls saturated with a copper salt. Copper sul-
phate is generally used to saturate the balls, but it is claimed
that this is soon converted into the chloride. The reactions
taking place in the tube are generally considered to be

%CuCl^z=z%CuCl'\-Cl^
2Cua+ (9, = 2CuO + Cl^
"iCuO -f 4:HCl = 2CuC/, + %Hfi

It is held by some, however, that the copper salt only
acts catalytically, and the reaction is direct between the
acid and the oxygen.

24 ALKALIES AND HYDROCHLORIC ACID § 30

I>ETAIL,8 OV THE PROCESS

•

24, HydrocMorlc Add. — The acid used for the Deacon
process must be of as uniform a composition as possible and
free from dust, sulphuric acid, and arsenic compounds, for
otherwise the contact substance deteriorates very rapidly.
The uniformity of composition is not hard to get when the
acid is liberated from its solutions. When it goes to the
decomposer direct from the salt-cake oven it is not so easy
to maintain a uniformity, for the acid is given off rapidly at
first and more slowly later. This difficulty is largely
avoided, however, by connecting several furnaces to each
decomposer, so that by charging the salt-cake furnaces in
rotation a nearly uniform flow of acid gas is obtained. Where
the acid is used direct from the salt-cake ovens, only the pan
acid is used; for this is much purer than that from the
roaster, and the acid from the latter can be condensed and
sold as acid or used in the Weldon process. At the present
time it is customary at many works to condense all the
hydrochloric acid produced and then liberate the gas from
its solution by running it into hot, concentrated sulphuric
acid and blowing a current of air through the mixture ; a
very pure hydrochloric-acid gas is thus obtained. This
method of purifying the hydrochloric acid was worked out
by Hasenclever, and has done much to make the Deacon
process a success; for this reason, the process is frequently
referred to as the I>eacon-Ha8enclever process.

Calcium chloride has been proposed for setting hydro-
chloric acid free from its solutions. It possesses no advan-
tage over sulphuric acid for this purpose, however, and the
latter is more generally used.

35, The hydrochloric acid is mixed with about an equal
volume of air, which furnishes the theoretical amount of
oxygen necessary to decompose it. Since, however, even in
the presence of a catalytic substance, the reaction is not com-
plete, an excess of air will drive the decomposition of the
hydrochloric acid farther. The disadvantage, however,
enters here, that the excess of air dilutes the already much

S 30 ALKALIES AND HYDROCHLORIC ACID 25

26 ALKALIES AND HYDROCHLORIC ACID § 30

diluted chlorine, so that it is better to allow a portion of the
acid to escape decomposition than to produce such dilute
chlorine. The mixture of air and hydrochloric acid must be
as dry as possible — the drier the better — before going to the
decomposer. It has been found in practical working, how-
ever, that very satisfactory results are obtained if the mix-
ture is cooled to 37° C. Gas saturated with moisture at that
temperature works in the hot decomposer nearly as well as
perfectly dry gas, and the cost of drying is saved.

36, A portion of the reactions in the decomposer absorbs
heat and a part evolves heat, but the sum total of these reac-
tions is an evolution of several calories of heat for each gram
molecule of hydrochloric acid oxidized. There is not enough
of this heat, however, to make up for loss through radia-
tion and also bring the gas mixture to the best tempera-
ture for the decomposition. It is, therefore, advisable to heat
the gas mixture to about 450** C. before it goes to the decom-
poser, as it is found that this is the best temperature for
decomposition.

!?7. The gas that issues from the decomposer consists
of a mixture of hydrochloric acid, chlorine, oxygen, nitro-
gen, and water vapor. Both the hydrochloric acid and the
water vapor must be removed if the chlorine is to be used
for bleach making. The gases, therefore, pass through a
cooling arrangement to condense the water as much as
possible, and with it the acid. It is then washed with water
and is finally passed through towers, down which sulphuric
acid is sprayed, to completely dry it.

38, Apparatus, — The apparatus for carrying out the
Deacon process is shown in Fig. 10. It consists of a cooling
and condensing arrangement for the gases as they come
from the salt-cake furnace or from the Hasenclever purifier.
This cooling and condensing apparatus consists usually of a
long, upright pipe A and a small coke or plate tower B,
The gas mixture goes to the heater C, which consists of a
series of pipes, up and down through which the gas must
pass. The pipes are enclosed and heated by the gases from

§ 30 ALKALIES AND HYDROCHLORIC ACID 27

a fire on the grate d. The gases having been heated to
about 450° C. pass into the decomposer E, Several forms of
this piece of apparatus have been proposed, but the one
here represented is the most satisfactory. It consists of a
large circular outer chamber, into which the mixture of air
and acid passes from the heater. Arranged inside of this
chamber, so that the gas must pass through them, are the
cylinders containing the catalytic material. The walls of
these cylinders are made similar to Venetian blinds, so that
the gas must take a downward course on entering, and
after traversing the filling, it takes an upward course on
leaving. The gases from the whole system collect in the
center and are drawn off by a pipe to the purifying appara-
tus. Each cylinder is arranged so that it can be cut out of
action when necessary for emptying and refilling, for the
catalytic material deteriorates slowly by use and must be
renewed about every 12 weeks. Frequently, in the style
of decomposer represented here, all the cylinders are kept in
continuous action, and when it is necessary to recharge
them the fresh material is charged at the top as rapidly as
the old is withdrawn at the bottom. For cooling and wash-
ing the gas that comes from the decomposer, a large number
of methods have been proposed, but the one illustrated at F
is probably the most efficient and at the same time the most
simple. It consists of upright pipes, which serve to cool
the gases, and end in troughs of water, which washes out
the hydrochloric acid. Finally, the gas is completely dried
by sulphuric acid in the towers (7. A suitable vacuum is
maintained in the whole apparatus by means of a pump
placed beyond G,

)89, Comparison of the Weldon and Deacon Proc-
esses.— It is difficult to say whether the Weldon or the Deacon
process leads in the production of chlorine at the present
time, and it is equally difficult to say which process is the
better, as this depends on general conditions.

In the old manganese-dioxide method, theoretically
50 per cent, of the chlorine of the acid was obtained free,

28 ALKALIES AND HYDROCHLORIC ACID § 30

but in practice not over 30 to 33 per cent, was realized. In
the Weldon process, only 40 per cent, of the chlorine of the
acid is theoretically available, but about 30 to 33 per cent.
is also obtained here and the manganese is recovered as
well. In both cases a strong chlorine is made. In the
Deacon process, 100 per cent, of the chlorine in the acid is
theoretically obtainable, and in practice 50 to 80 per cent. ;
the rest is recovered as acid to be used over. The chlorine
is much diluted, however, only averaging 7 to 10 per cent,
chlorine, so that it is not so suitable for as many purposes
as the stronger gas obtained from the other methods.

THE NITRIC-ACTD CHIX>RINE PROCESS

30. A number of processes have been proposed that
involve the oxidation of hydrochloric acid by means of
nitric acid, according to the reaction

ZHCl + HNO, = ^Hfi + NOCl + C/,

This gives two-thirds of the chlorine in a free state and
leaves one-third combined in nitrosyl chloride. It is then
necessary to set the chlorine free from this compound and
regain the nitric acid by oxidation. This is accomplished
by treating the nitrosyl chloride with concentrated sulphuric
acid and then with air and steam. All the diflEerent processes
belonging to this class employ this reaction, and only differ
in the methods of mixing the materials so as to obtain the
best results. They all give a high yield of very concentrated
chlorine; there is very little loss of hydrochloric acid, and
95 per cent, of the nitric acid can be recovered and returned
to the process. On the other hand, the handling and con-
centrating of such large quantities of acid as are required
are difficult and dangerous, and the wear and tear on the
apparatus is very considerable.

31. Where chlorine is used in large quantities it is some-
times made on the spot, either directly from salt or from
hydrochloric acid. The use of salt is, however, almost

§ 30 ALKALIES AND HYDROCHLORIC ACID 29

obsolete, and the carrying of hydrochloric acid is incon-
venient and somewhat dangerous. For this reason it can
rarely be economically made at any place far removed from
alkali works. On the other hand, the chlorine gas is
bulky and must be converted into some compact form for
shipment.

32, lilquld Chlorine. — Chlorine is a gas that is com-
paratively easily liquefied, for it becomes liquid when cooled
to —34*' C. at the ordinary atmospheric pressure, or when
subjected to a pressure of 6 atmospheres at the ordinary
temperature. It is such a corrosive substance, however,
that until recently it was not considered possible to find
pumps to work it, or tanks to hold it when it was com-
pressed. The pumps used in compressing chlorine consist,
for the most part, of a plunger that works in petroleum
and forces the petroleum against a column of sulphuric
acid. The chlorine collects over the acid, and when the
acid is raised the chlorine is forced into a tank and com-
pressed. Moist chlorine acts very strongly on iron at the
ordinary temperature; but when perfectly dry, chlorine has
practically no action on iron, and iron tanks can be safely
used for storing and shipping it when in the liquid form.
One volume of liquid chlorine is equal to 400 volumes of
chlorine gas at ordinary conditions of temperature and
pressure.

BLEACHING POWDER

33. When chlorine is passed over dry, slaked lime a com-
pound is formed that again gives up the chlorine when treated
with an acid. This compound was at first supposed to be
calcium hypochlorite Ca{OCl)^ and was called chloride of
lime. It is now more commonly known as bleachijig
powder. Bleaching powder only yields 100 volumes of
chlorine for each volume of the substance and requires acid
to set it free. It is, nevertheless, a most convenient means
for the transportation and storing of chlorine and is almost
universally used.

30 ALKALIES AND HYDROCHLORIC ACID § 30

34, Xdme. — The lime used for making bleaching powder
should be very pure and well burned. Impurities are bad
in various ways, for in addition to making it impossible to
make a strong bleach, if the lime does not contain a high
percentage of calcium oxide, clay and similar substances
cause the bleach solutions to settle badly. Iron and man-
ganese cause a colored bleach, which does not sell well, and
these substances cause a more rapid decomposition of the
bleach than would otherwise occur. A limestone of as great
purity as possible having, therefore, been selected, it is
burned in such a manner as to avoid having the ashes of the
fuel mix with the lime. A reverberatory furnace is fre-
quently used for this purpose. The carefully burned quick-
lime is slaked by sprinkling with water; -and as an excess of
water cannot be used, it is better to let the lime lie for 2 or
3 days to allow it to slake well through before using. Per-
fectly dry, slaked lime does not work well with chlorine and,
on the other hand, too great an excess of water must be
avoided or the lime will cake together and not chlorinate
through. Theoretically, calcium oxide requires 32 percent,
of its weight of water to convert it into the hydrate, and
from 2 to 4 per cent, of water in addition to this, depend-
ing on the dehydration of the chlorine, is generally
used. After slaking thoroughly, the lime is sifted through
a sieve having from 12 to 25 holes to the linear- inch.
The finer the division of the lime, the better it absorbs
the chlorine. It is now ready to spread in the absorption
chambers,

35, Absorption Chambers, — The chambers for absorb-
ing the chlorine are commonly large rooms made of brick or
stone laid in asphalt cement; though they are sometimes
made of lead, which is probably the best material and is not
much more expensive than the other. The floors are either
of asphalt or lead. The lime in the chambers must be turned
over when the layer is thick, so that the chambers must be
high enough for a man to stand upright in while turning
and removing the material. An ordinary chamber is about

§ 30 ALKALIES AND HYDROCHLORIC ACID 31

100 feet long, 30 feet wide, and 6^ feet high. It is usually-
estimated that 200 square feet floor space is required per
ton of bleach per week. The slaked lime is spread on the
floor in a layer from 2 to 4 inches thick and is furrowed by a
rake to give a large absorbing surface. The gas passes
into the chamber at the top of one end and out of the top
of the opposite end. The chlorine, being heavy, settles to
the bottom of the chamber and is very rapidly absorbed at
first and then more slowly, as the lime becomes more nearly
saturated. In the case of single chambers, when the absorp-
tion becomes too slow, the gas is shut off and, after freeing
the chamber of chlorine, men go in and turn and relevel the
lime. In the more modern works, where three or more
chambers are worked together, the turning can be avoided,
for the strong gas goes into the most nearly finished cham-
ber and then to fresher lime, so that the chlorine does not
escape. When the layer of lime is not over 2 inches thick
the operation will usually be finished without turning th^
material ; when the layer is over 2 inches, the material must
usually be turned. A second passing of the gas will usually
bring the available chlorine in the bleach to 36 to 38 per cent.,
and that is sufficient. If this is not obtained, the material
must be turned a second time and then treated with gas
again. If this does not bring the bleach to the desired
strength, it must be packed and sold for what it will bring,
for further treatment with chlorine will only result in the
decomposition of the bleach already formed.

36, Chlorine. — The chlorine must be free from carbon
dioxide and hydrochloric acid and as free from water as pos-
sible. The stronger the chlorine the better, and very dilute
chlorine, such as comes from the Deacon method, cannot be
used in this form of apparatus. The chlorine must be
introduced into the chamber very slowly, so as to avoid a
rise in temperature, for if the temperature is too high, chlo-
rates will form and the bleach decompose, giving oxygen.
On no account should the temperature go above 40° or 45° C. ,
and a lower temperature is better.

32 ALKALIES AND HYDROCHLORIC ACID § 30

37, The opening of the chamber to turn or remove the
bleach is disagreeable, for the chlorine escapes into the air.
This is obviated somewhat by letting the chambers stand for
some time before opening, or, better, by sprinkling a little
fine dust of calcium hydrate in from the top. At best
there is a great deal of hard and disagreeable work connected
with the process, and the plant covers a large area. The
attempt has been made to do away with these difficulties by
stirring the lime mechanically while it is being chlorinated.
By this means the lime is chlorinated rapidly and discharged
into the barrels without much hand labor. The great dis-
advantage exists that, by such rapid absorption of the chlo-
rine, the temperature gets too high. This has been somewhat
obviated lately by cooling the apparatus from the outside.

38, As already mentioned, the apparatus that is suitable
for strong chlorine cannot be used for the more dilute chlo-
rine obtained in the Deacon method, for the absorption is
too slow with such weak gas. Deacon avoided this difficulty
by using large stone chambers in which shelves were placed
close together. On the shelves the finely powdered slaked
lime was spread in layers not over | inch thick and the chlo-
rine passed downwards over these shelves. This arrange-
ment works very well, but it requires very large chambers.
For each ton of bleach produced in a week a shelf space of
1,373 square feet is necessary. With the dilute chlorine,
the absorption is not so rapid and the mechanical chlorina-
ting apparatus can be used to good advantage.

39, Properties of BleaeUng: Powder. — The chloride
of lime should be a white powder or in lumps that will easily
break. It is acted on by the carbon dioxide of the air, and
so loses strength if left open ; even when protected from the
air it slowly loses strength, especially when it is jarred as in
transport. It has a peculiar odor, probably due to chlo-
rine. It is usually packed tightly in barrels to exclude air
and moisture, and these should be kept out of the sun as far
as possible. The bleach loses about 1 per cent, of chlorine in
packing (probably chlorine that is mechanically held in the

§ 30 ALKALIES AND HYDROCHLORIC ACID 33

bleach) and then should have from 33 to 38 per cent, of
available chlorine at the works. When bleach is imported
into this country, it rarely contains over 32 or 33 per cent^
of available chlorine, the rest being lost in transportation.

40. Composition of BleacMngr Powder. — When

bleaching powder was first made, it was considered to be

calcium hypochlorite Ca{OCl)^. It was then shown that this

was improbable and that certain considerations seemed to

lead to the view that it was a mixture of calcium chloride

and hypochlorite CaCl^ + Ca{OCl)^, There are, however,

several reasons for thinking that this formula is incorrect.

Among others, it might be mentioned that if it contained

calcium chloride it should be deliquescent, but bleach is not ;

calcium chloride is soluble in alcohol, but it cannot be

extracted from bleach by this means. Lunge has proposed

CI
the formula Ca <^ ^^. for the substance, and has so well

supported this view by experiment, that it is generally
accepted as correct. When bleach is dissolved in water, it
breaks up into calcium chloride and hypochlorite.

41. Valuation of Bleach. — The only constituent that
bleaching powder contains that is of value is the chlorine
that can be utilized for bleaching purposes. The amount of
the available chlorine is determined by analysis, and in most
countries, outside of France, the value of the bleaching
powder is expressed in terms of the percentage of the avail-
able chlorine contained, as shown by analysis. For example,
a 32-per-cent. bleach means that the bleach under considera-
tion contains 32 per cent, of chlorine that is available for
bleaching purposes. In France, and to some extent outside
of that country, the strength of the bleach is expressed in
Gay-Lussac degrees — that is, the number of cubic centimeters
of chlorine gas, reduced to the standard conditions of 0° C.
temperature and 760 millimeters of mercury pressure that
1 gram of the bleaching powder will yield. If we remember
that 1 gram of chlorine under standard conditions occu-
pies 314.7 cubic centimeters, it is easy to calculate the

34 ALKALIES AND HYDROCHLORIC ACID § 30

Gay-Lussac degrees from the percentage in the composition.
For example, if we have 32 per cent, of available chlorine in
a sample of bleach, each gram of the bleach contains .32 gram
of available chlorine and will yield 314.7 x .32 = J 00. 7 cubic

TABLE I

Gay-
Lussac

English

Gay-
Lussac

English

Gay-
Lussac

English

Degrees

20.02

27.01

107

34.00

20.34

27.33

108

34.32

20.65

27.65

109

34.64

20.97

27.96

34.95

21.29

28.28

III

35.27

21.61

28.60

112

35.59

21.93

28.92

35.91

22.24

29.23

114

36.22

22.56

29.55

. "5

36.54

22.88

29.87

116

36.86

23.20

30.19

117

37.18

23.51

30.41

118

37.50

23.83

30.82

119

37.81

24.15

31.14

120

38.13

24.47

31.46

121

38.45

24.79

100

31.78

122

38.77

25.10

lOI

32.09

123

39.08

25.42

102

32.41

124

39.40

25.74

103

32.73

125

39.72

26.06

104

33.05

126

40.04

26.37

105

33-3^

127

40.36

26.69

106

33.68

128

40.67

centimeters of chlorine under standard conditions, or it is
100.7° Gay-Lussac, strong. These are sometimes called
French degrees^ and the percentage of available chlorine in
the bleach is frequently called English degrees. Table I

§ 30 ALKALIES AND HYDROCHLORIC ACID 35

shows at once the relation between the Gray-Lussac degrees
and the English degrees.

43. Uses. — Bleaching powder is mostly used for bleaching
vegetable fibers. The fiber to be bleached is first saturated
with the bleach in clear solution, it is then ** soured "by pass-
ing it through dilute acid, and is finally washed. Since the
bleaching powder must be dissolved, it would seem that it
might better be made direct in solution, as was done in the
early days of the industry. The solution of bleaching powder
does not keep well, however, and the large amount of water
makes it inconvenient and expensive to transport. The
liquid bleach is, therefore, only made in the few cases where
the bleaching establishment is near an alkali works. In
making liquid bleach, the chlorine is not passed through the
milk of lime, for this would put too much pressure on the
chlorine stills, but goes over the surface of the liquid and is
so absorbed.

43. Eau de Javel. — The first bleach that was made
was prepared by passing chlorine into a solution of potas-
sium carbonate (crude potash). As the works were situated
at Javel, near Paris, it took its name from that place. A
little later sodium carbonate was substituted for the potash,
and the solution made from this substance became known as
Eau de Labarraque, This latter substance is still some-
times made and used for certain purposes. When the chlo-
rine is passed over a sodium-carbonate solution, the first
action is to convert the carbonate into the bicarbonate and
form hypochlorous acid, according to the reaction

Na^CO^ -f a, -f Hfi = NaCl-^' HNaCO, -f HCIO

If the chlorine is passed long enough, the bicarbonate is
decomposed and the carbon dioxide evolved. This reac-
tion is

NaHCO^^Cl^^NaCl'^CO^'^HClO

In this case, however, chlorate is likely to be formed.
Another class of liquor, which is more stable than the above,

36 ALKALIES AND HYDROCHLORIC ACID § 30

is made by passing chlorine over caustic soda. The solution
must be left slightly alkaline and kept cool to prevent the
formation of the chlorate. The reaction then is

%NaOH + a, = NaCl + NaClO + Hfi

By this means we have a fairly stable solution of bleach-
ing material. Until recently it was not considered possible
to make this bleach solution stronger than 15 per cent, of
available chlorine, and that strength kept badly. It has
been found, however, that this instability is caused by the
presence of sodium ferrate, which acts catalytically and
causes the solution to decompose. When the sodium hydrate
is carefully purified from iron, solutions of the hypochlorite
containing as high as 50 per cent, of available chlorine can be
made, and solutions with 35 per cent, of available chlorine
are quite stable. Solutions with 20 per cent, of available
chlorine can be kept for weeks with practically no change.
The solution must be kept slightly alkaline, however, or the
hypochlorite will change over into the chlorate.

44. With aid of the bleach liquors so far spoken of,
it is necessary to use acid to get the bleach effect, and then
it is necessary to wash thoroughly. Sometimes this is dis-
advantageous, and other hypochlorites are made that
decompose more readily on the fiber and so do not need
acid. These are practically all made from the calcium
hypochlorite. The aluminum bleach is the most important
of these, and its method of preparation is typical of the
method used in the preparation of all the others.

The aluminum bleach consists of a solution of a mixture
of aluminum chloride and hypochlorite that is made by
treating a solution of calcium bleach with aluminum sul-
phate ; the calcium sulphate separates out and the aluminum
compounds are left in solution. The aluminum hypochlo-
rite is very unstable and is only made as needed. It is so
very unstable that it decomposes on the fiber without the
use of acid, and the aluminum compound left is antiseptic,
so that it not only does not need to be washed out, but in

§ 30 ALKALIES AND HYDROCHLORIC ACID 37

many cases it is a decided advantage to leave it on the
bleached material. For example, when used to bleach paper

stock the aluminum chloride prevents fermentation when
the stock is stored.

POTASSrUM CHIiORATB

45. There are at present two general methods for
making potassium chlorate, the electrolytic, which will be
discussed in its proper place, and the chemical. The most
generally used chemical process consists in making calcium
chlorate and converting this into potassium chlorate by
adding potassium chloride and allowing the less soluble
potassium chlorate to crystallize out. The calcium chlorate
is made by absorbing chlorine in milk of lime ; so that prob-
ably calcium hypochlorite is first formed and this is trans-
formed into the chlorate. The reactions taking place are
doubtless

%Ca{OH)^ + %Cl, = CaC/, + Ca{OCl\ + 2//,6^
dCa{OCl), = Ca{ClO,), + %CaCl,
or QCa{Off), + 667, = 6CaC7, + Ca{C/0,), + 6//,0
and Ca{aO,), + 2KC/ = CaC/^ + 'Z/CC/0^

A greater saving is made in this way than would be made
by starting with caustic potash instead of caustic lime.

RAW MATERIAL-S

46. liiine. — The lime used for this process should be the
very best and as free from impurities as possible. It is
usually burned in a reverberatory furnace. The thor-
oughly burned lime is slaked, made into milk of lime,
and strained before it goes to the absorbers. It should
then be used without delay, as otherwise calcium carbonate
will form, and this leads to a loss of chlorine.

47. CMorine. — Chlorine made by either the Weldon or
the Deacon process can be used, and generally no attempt

38 ALKALIES AND HYDROCHLORIC ACID § 30

is made to remove the water and carbon dioxide. The
hydrochloric acid is only removed when it occurs in such
large quantities as in the Deacon process. Chlorine made
by Weldon's process is much preferred, as it is stronger and
so gives better absorption.

48. Potassium Chloride. — The potassium chloride used
is almost entirely imported from Germany and contains
from 90 to 93 per cent, of potassium chloride. The other
constituents are mostly soluble and do but little harm.
The following analysis gives a fair idea of the average
composition of commercial potassium chloride, so-called
muriates,

H^O 4.50^ Na^SO^ 30^

Organic 05^ CaCl^ 25j^

Insol. and Fefi^ .15^ MgCl^ 50^

Alfi^ 47^ NaCl %%hi

A/,{SO,), 20j< KCl 92.00^

49. Water. — The water, especially that used for crys-
tallization, must be pure. Suspended matter tends to pre-
vent the formation of crystals and leaves them impure
when formed. The presence of sulphides leads to the for-
mation of lead sulphide, for there is usually lead in the final
liquor from the lead crystallization pans. Sulphates are
liable to be reduced by organic matter and so lead to the
presence of sulphides, so they must be excluded; for the
lead sulphide would make the crystals dark colored and
spoil their sale. Iron and carbonates are also objectionable,
but are not so bad as the other substances.

APPARATUS AND PROCESS

60. Aljsorbers. — In making the calcium chlorate, the
chlorine must be passed over the surface of the milk of
lime. The absorption of the chlorine by this material takes
place in large, flat, quadrangular tanks, which are built of
slabs of sandstone. Where the sandstone slabs come together,

§ 30 ALKALIES AND HYDROCHLORIC ACID 39

they are grooved out and a thick rubber cord is introduced.
The whole is then tightly fastened together with iron tie-
rods placed around the outside. In order that the absorp-
tion may take place more readily, each tank is fitted with an
agitator that stirs and splashes the milk of lime so that an
intimate mixture of it and the chlorine takes place. These
agitators pass into the tanks through hydraulic lutes ; the
manholes in the tanks are also provided with hydraulic
lutes, so that the tanks are closed tightly when in operation.

The absorbers are usually set up in series of from three
to five, so that the liquor can flow from one to the next and
the chlorine enters the absorber that is most nearly finished
and leaves the one newly charged. The gas that leaves the
last, absorber is nearly free from chlorine, but is finally run
through a tower, down which milk of lime is flowing, in
order to remove the last trace of chlorine before the gas
escapes into the air.

In carrying out the operation, the lowest absorber is
emptied when the absorption is complete and the contents
of each absorber run into the next lower one. The upper
absorber is then charged with milk of lime of 1.085 or
1.100 sp. gr. (that is, about 113 grams CaO per liter). The
absorber should not be charged over two-thirds full, for
there is danger that it will foam over at some stage of the
absorption. Chlorine is now passed into the lowest absorber
and continued until all of the lime is converted into cal-
cium chloride and calcium chlorate. As the chlorine is
absorbed, the temperature of the absorbing liquid grad-
ually rises and must be carefully watched. The tem-
perature should not be allowed to exceed 55° C. , or the
yield of chlorate will suffer in consequence. The tem-
perature can be very easily regulated by regulating the flow
of chlorine. The charge requires from 12 to 30 hours from
the time it is first run in until it is finished. The time
depends on the size of the absorbers and the strength of
the chlorine gas and the milk of lime. Slow absorption,
using weak solutions, gives the best results from a chemical
point of view, but, on the other hand, more concentrated

40 ALKALIES AND HYDROCHLORIC ACID § 30

solutions and quick absorption save time and fuel, so that
a balance must be struck for each locality, depending on
the price of coal.

The end of the reaction in the absorber is shown by the
appearance of a pink color, due to the formation of calcium
manganate from manganese in the lime or carried over with
the chlorine. Another rapid test consists in filtering oflf a
little of the solution and adding dilute hydrochloric acid to
it. An effervescence, or evolution of chlorine, shows that
the solution still contains calcium hypochlorite and that the
operation is incomplete.

61. Settling Pans. — When the absorption is completed,
the finished liquor is run into large iron pans, where it is left
for from 3 to 10 hours for the insoluble matter, such as sand,
calcium carbonate, etc., to settle out. The capacity of the
settling pans must at least equal the capacity of the absorbers,
for on account of the sand, etc. , that settle in these pans their
actual capacity is frequently much less than their nominal.
When the liquor has settled thoroughly, it is pumped by
means of force pumps having gun-metal barrels to a higher
level, in order that it may then run by gravity through the
rest of the operations. The best suction pipe for the pump
is a short rubber hose, which can be moved so as to suck
the liquor close to the mud, without getting part of the
latter into the concentrating pots.

The mud is allowed to accumulate in the pans until they
are nearly half full, it is then washed two or three times; the
wash water is used in making milk of lime, while the mud is
thrown out.

63. Concentrating: Pots. — The liquor from the settling
pans is carefully gauged and a sample sent to the labora-
tory for analysis. Meanwhile, the liquor goes to the con-
centrating pans, which are best made of cast iron and are
similar in size and shape to those used in making caustic
soda (see Alkalies and Hydrochloric Acid, Part 1), and is
here warmed. By this time the analysis of the liquor should
be made and the amount of potassium chloride necessary to

§ 30 ALKALIES AND HYDROCHLORIC ACID 41

convert the calcium chlorate into potassium chlorate is cal-
culated. This amount, plus about 1^ per cent., is then
added and the whole concentrated to about 1.31 sp. gr.
(taken hot). In winter, a slightly lower specific gravity
will answer.

53« First Ciystalllzlngr Pans. — The concentrated liquor
is now baled into the crystallizing pans. These are usually
U-shaped and are set into brickwork a slight distance above
a cement floor. These pans are built of iron and should be
of such a size that the contents of a pot just fills a certain
number of them. The room in which these pans are set
should have a cement floor that slopes towards a catch basin.
The pans are left for 9 or 14 days, depending on the time
of year, to crystallize the liquor.

The crystals are filtered off by means of a centrifugal
machine, thoroughly washed with water to remove the cal-
cium chloride and iron, and then recrystallized.

The mother liquor, which is mainly calcium chloride, con-
tains from 10 to 35 grams of potassium chlorate per liter,
and is cooled to— 10**C. by artificial means. In this way
the amount of potassium chlorate is reduced to about
3 grams per liter.

54. Recrystallizatlon. — The crystals obtained by the
first recrystallization always contain considerable impurities
and are therefore placed in a large, lead-lined, iron cylinder,
water is added and steam blown in until the solution has a
strength of 1.10 to 1.11 sp. gr. (taken hot). This apparatus
is placed high enough so that the solution can be drawn
direct to the crystallizing pans through 3-inch, steam-heated
steel pipes. These operations are carried out in a separate
building and with all possible cleanliness. The crystallizing
vats are usually of iron and are lead-lined; a convenient
size is 5 feet by 4 feet, and 3 feet deep. They should be
raised a little above the cement floor, so that leaks can
be detected; and the floor should slope to a catch basin,
to avoid loss oi^ the liquor accidentally spilled. From 7 to
10 days are allowed for the crystals to separate out ; they are

42 ALKALIES AND HYDROCHLORIC ACID § 30

then filtered off in a centrifugal machine and washed until
not over .05 per cent, of chlorides is shown by testing.

The mother liquor is used for dissolving fresh crystals
until it reaches a specific gravity of about 1.08, when it is
too impure and is stored until enough is obtained, when it is
boiled down and crystallized for crude crystals. The mother
liquor from these is run into the ordinary concentrating pots.

66. Drying tlie Crystals. — The thin transparent crys-
tals are thoroughly drained and then put on to the drying
table, which consists of a table of boiler iron having an up-
turned rim and covered with lead. It is heated by steam.

66. Grinding tJie Crystals. — For many purposes the
dry crystals can be marketed direct ; but for others, they
must be ground to a fine powder. This is a very danger-
ous business and must be performed with the greatest care.
The engine for driving the mill is situated outside of the
building and all inflammable material is excluded so far as
possible. The crystals are ground between small stones
(about 26 inches in diameter), of which only the top one
revolves. The crystals are fed in at the center of the top
stone through a hopper and are best ground warm from the
drying table, as in this way the mill clogs less. The ground
crystals are then sifted through mechanically rocked sieves
and the fine powder is packed.

OTHER CHLORATES

67. Sodium Chlorate. — Sodium chlorate is more solu-
ble than the potassium salt, and for this reason is better
suited for many purposes. It is, however, for the same rea-
son, not so easy to make, for it cannot be readily separated
from the other substances in solution. It can be made from
the calcium-chlorate solution by evaporating it to 1.5 sp. gr.,
and then cooling to 10° or 12° C. The calcium chloride is
crystallized out until there is only 1.2 molecules of calcium
chloride to 1 molecule of calcium chlorate. By then adding

g 30 ALKALIES AND HYDROCHLORIC ACID 43

sodium sulphate and a little sodium carbonate, all the cal<
cium is precipitated and sodium chloride and chlorate are
left in solution; then by boiling down, the salt is separated
out and the chlorate is left -alone in solution. The solution
is then run off and cooled, when most of the sodium-chlorate
crystallizes out free from salt.

Hargreaves makes sodium chlorate by the direct action
of chlorine on crystalline sodium carbonate and systematic
leaching, so as to
dissolve out the
soluble chlorate
and leave the less
soluble salt behind.
He places the crys-
tallized sodium
carbonate in the
tower b. Fig. 11,
which is supported
on the grate c c;
the chlorine enters
at d and, passing
upwards, is a b-
sorbed. Liquor
from the tank e
slowly trickles
down over the
charge and is run
off through /,
where it goes into
the sieve, which
holds back any
solid material, and
the liquid runs
through into the

cistern, from ""' "

which it is pumped back to e until it is saturated with
sodium chlorate. It is then run off to pans and cryS'
tallized.

44 ALKALIES AND HYDROCHLORIC ACID § 30

58. Barium chlorate and other chlorates can be made
from sodium chlorate by mixing the chloride of the metal
whose chlorate is wanted, evaporating down, and fishing out
the sodium chloride. The metallic chlorate then separates
out in cooling. These chlorates may also be made in a simi-
lar manner to the methods given above for the making of
sodium chlorate.

ELECTROLYTIC METHODS

GENERAL. PRlNCIPIiES

THE CURRENT

69. Sources of Current. — There are three methods
for producing a continuous flow of electricity; i. e., the
voltaic cell in some one of its various forms, the dynamo, and
the thermopile. Of these, the voltaic cell is too expensive
to be used as a source of electricity for electrolytic work on
a commercial scale, for its action depends on the dissolving
of expensive materials. In the thermopile we obtain a
flow of electricity by heating the junction of two metals, and
thus converting heat directly into electricity. This method
is, however, wasteful of heat and is also too expensive for
commercial use. The dynamo depends for its action on the
rotation of a coil of wire in the field of force of a magnet,
and as the coil can be rotated by means of a steam engine,
or, better still, by water-power, it furnishes the most eco-
nomical source of electricity at present known. The
dynamo current is generally used direct from the machine,
but it may be stored for future use by means of a special
form of battery, called a storage battery. The storage
battery also has the advantages that it can be transported and
that it will yield a uniform current. Any voltaic cell
which after being used can be returned to its original con-
dition by the passage of an electric current in the opposite

§ 30 ALKALIES AND HYDROCHLORIC ACID 45

direction is, in the perfect sense of the word, a storage bat-
tery. Only one form of battery has, however, proved itself
useful for practical purposes. This consists of a plate of
lead coated with lead peroxide on both sides and a plate of
spongy lead dipped in a solution of sulphuric acid. If,
under these conditions, the two lead plates are joined by
a wire, the lead becomes transformed into lead sulphate
and hydrogen separates on the lead-peroxide plate. Here
the hydrogen is oxidized to water and the lead peroxide is
reduced to lead oxide, which also goes over into lead sul-
phate. Now, when a current is passed into the cell, the
reverse operations go on and the cell is returned to its orig-
inal condition.

For convenience of reference, we will refer to the stor-
age battery as our source of current, although it must be
borne in mind that all the statements made will hold
equally well for the current from any other source, at least
so long as it is not an alternating current.

60. Just as when two unconnected dishes of water are
placed on different levels there is a latent power in the
water in the higher dish that gives it a tendency to flow
into the lower one, which it does when they are connected
by an open tube; so the plates of the storage battery are
latent so long as they are not connected, but as soon as
they are joined by a wire a current flows from the plate
that corresponds to the higher dish into the plate that cor-
responds to the lower dish of water. In the case of the
water, we say it has a **head " of a certain amount, meas-
ured by the difference of level of the two dishes ; in the
case of the electricity, we call it a difference of potential
and measure it in a unit called a volt. This difference of
potential of the plates of a cell is called the electromotive
force of the cell. The water in flowing through the tube
is retarded by the friction in the tube, and therefore does
not reach the lower level with as much force as would
otherwise be the case. The electricity is resisted by the
conductor, and this resistance is measured in ohms. The

46 ALKALIES AND HYDROCHLORIC ACID § 30

quantity of electricity, corresponding to the quantity of
water, is measured in coulombs and its rate of flow in
amperes.

Or, since electricity is, unlike water, an imponderable
substance, or rather, a manifestation of energy, perhaps its
analogy to heat is a better one than the above. In this
case, the difference of potential, or electromotive force, cor-
responds to the difference of temperature of two points, the
resistance of the conductor corresponds to the non-conduc-
tivity of the connecting medium for heat, and the quantity
of current corresponds to the quantity of heat, in calories,
that passes from the point of higher temperature to that of
lower.

61. Units of Measurement. — Just as in measuring dis-
tance, a certain distance, as the foot, or meter, is arbi-
trarily selected as a unit to express the distance, or in
measuring differences of temperature some definite differ-
ence of temperature, as a degree, is selected to express the
difference of temperature; so in electrical measurements,
a unit has been carefully selected in which to express the
amount of the various values in which we deal.

The unit of resistance, the ohm^ is the resistance at 0** C.
of a column of mercury 1 square millimeter in section and
1.0626 meters long. The unit quantity of electricity,
the coulomb^ is the quantity of electricity that will deposit
1.118 milligrams of silver from the solution of a silver
salt under suitable conditions. The unit of difference
of potential, or electromotive force, the volt^ is the differ-
ence of potential that will send 1 coulomb per second
through a resistance of 1 ohm. The unit of the rate of
flow of a current, the ampere^ is the rate of flow that will
carry 1 coulomb past a point on the conductor each second.
The unit of electrical power, the watt^ is the product of the
volt and ampere and is equivalent to '^\^ horsepower, or, in
other words, 746 watts equal 1 horsepower. The unit of
electrical energy, the volt coulomb^ or joule^ is the product
of the volt and coulomb and is equivalent to .24 calory.

§ 30 ALKALIES AND HYDROCHLORIC ACID 47

The current density is measured by the number of amperes
entering or leaving the solution per unit surface of the
electrodes. It is usually expressed in amperes per square
decimeter, although other units of surface are also some-
times used, as the square meter or square foot, etc.

MSASUBEMENTS

63. Resistance. — Electrical resistance may be measured
by an apparatus called a Wheatstone bridge. A bridge
when completed, ready for taking measurements, consists of
three main parts: (1) An adjustable resistance box con-
taining a number of coils, the exact resistance of each coil
being known ; (2) a galvanometer for detecting small cur-
rents; and (3) a bat'tery of several cells. The coils of the
resistance box are divided into three groups, two of which
are called proportional or balance arms, and the third is
known as the adjustable arm. Each proportional arm is
composed of three and sometimes four coils of 1, 10, 100,
and 1,000 ohms resistance, respectively. The adjustable
arm contains a large number of coils ranging from .1 ohm
up to 10,000 ohms.

The operation of the bridge depends upon the principle
of the relative difference of potential between two points in
a divided circuit
of two branches.
The electrical con-
nections of the
bridge are shown
in the diagram.
Fig. 12. M rep-
resents the resist-
ance of one of the
balance arms,
which will be
termed for convenience the upper balance arm; N rep-
resents the resistance of the other balance arm, which

MWr

B
PIO. 12

48 ALKALIES AND HYDROCHLORIC ACID § 30

will be termed the lower balance arm ; P represents the
resistance of the adjustable arm; and X represents an
unknown resistance, the value of which is to be deter-
mined. One terminal of the detecting galvanometer G is
connected at f, the junction of the upper balance arm and
the unknown resistance; the other terminal is connected
at d^ the junction of the lower balance arm and the adjust-
able arm. One pole of the battery is connected at a^ the
junction of the two balance arms; the other pole at b^ the
junction of the adjustable resistance and the unknown
resistance. The current from the battery divides at a^
part of it flowing through resistances M and X^ and the
rest through iVand P. When the resistances J/, A^, P, and

M X
X fulfil the proportion -v^ = yj, then the two points c and d

will have the same potential, and no current will flow
through the galvanometer G. Since the resistances of J/,
N^ and P are known, the resistance of X will be given

by the fundamental equation Jf=-jrvX Z', when the arms

are so adjusted as to cause no deflection of the galvanom-
eter. For example, suppose that the two ends of a copper
wire are connected to the terminals b and r, and after
adjusting the resistance in the arm so that the galvanom-
eter shows no deflection, the resistances of the different
arms read as follows: J/= 1 ohm, 7^=100 ohms, and
P= 112 ohms. Then, substituting these values in the
fundamental equation gives

M 1

X--^y. P= — X 112 = 1.12 ohms.

The coils of resistance can be bought already put up in
boxes and standardized so that it is frequently more con-
venient to buy them in that way than to make them. They
are called resistance boxes. In these resistance boxes, the
ends of the wire of each spool are fastened to metal pieces a.
Fig. 13, so arranged that the metal pieces can be con-
nected by a metal pin b. When the pin b is in place, the

§ 30 ALKALIES AND HYDROCHLORIC ACID 49

current can flow from one plate a to the next through
the pin, and there is practically no resistance. When
the pin is removed, however, the
current must flow through the wire,
and the resistance is introduced.

Just as a certain resistance is found
when it is attempted to pass an elec-
tric current through a wire, so is a
resistance met when a solution is
used as a conductor. The deter-
mination of the amount of this re-
sistance is a matter of importance.

FIG. 13

63. Conductivity of 8oIutions.

Although it is customary to speak of
the resistance of a wire, we sometimes
hear the conductivity spoken of, and in the case of solutions,
it is much more common to speak of the conductivity than
of the resistance. The unit of conductivity, which has no
special name, is the conductivity of a body that, for 1 centi-
meter length and 1 square centimeter base, has a resist-
ance of 1 ohm. The specific conductivity of a solution is
the conductivity of a centimeter cube of the solution. The
conductivity of solutions is, however, expressed as the
equivalent [conductivity of the solution ; this is the specific
conductivity multiplied by the number of equivalent weights
in grams of the dissolved substance in 1 cubic centimeter
of the solution. By the term equivalent weight we mean
the molecular weight divided by the number of valences
represented in the metal part of the salt ; or in the case of
acids, by the number of acid-hydrogen atoms. For example,

^^^^\ HNO,, CH.COOH, ^l-\ ^^, etc., if the

formulas are expressed in terms of the atomic weights, are
equivalent weights.

64. Effect of Temperature. — The conductivity of
solutions increases very rapidly with a rise of temperature.

60 ALKALIES AND HYDROCHLORIC ACID § 30

The amount of the increase varies for different solutions, but
it averages about 2 per cent, of the conductivity for each
degree rise of temperature; of course, a fall of temperature
gives the reverse effect. It is therefore very necessary to
keep the solution at a definite temperature while making
the conductivity measurements. For this reason the vessel
containing the solution is kept in a constant-temperature
bath during the whole time of the measurement.

65. Constant-Temperature Bath. — A suitable con-
stant-temperature bath for technical work is made by
wrapping a wooden pail in felt, as by this means water at
nearly the temperature of the room can be kept at a constant
temperature for a long time. (Most determinations are
made at either 18° or 26° C.) With an arrangement of this
kind the desired temperature can be obtained by mixing
hot and cold water, and the temperature watched by a ther-
mometer hanging in the water. When higher temperatures
are to be used or a number of determinations are to be
made at one time, more elaborate apparatus can be arranged,
with stirrers and automatic temperature regulators.

66. Conductivity Vessel. — The form of the conduc-
tivity vessel will ,be different, depending on the conductivity

of the solution. For solutions of low
conductivity, as the organic acids, am-
monia, etc., a resistance vessel is neces-
sary with broad electrodes placed close
together; for better conducting solu-
tions, as inorganic acids, salts, and caus-
tic alkalies, a small surface of electrodes
with a rather long and small connecting
tube is more suitable.

For the first class of solutions, such a
vessel as is shown in Fig. 14 is the most
suitable. It consists of a cylindrical
cflass vessel 5, fitted with a hard-rubber

Pig 14

cap b having three holes, one for a
pipette, when it is necessary to introduce or remove liquid.

[

§ 30 ALKALIES AND HYDROCHLORIC ACID 61

and the other two for the electrodes. The electrodes con-
sist of two circular platinum disks r, c fastened by means
of heavy platinum wire into the capillary glass tubes d^ d.
The capillary tubes are filled with mercury, which makes
a connection between the ends of the platinum wires from
r, c and the copper wires that lead to the other connections.
The glass tubes d, d are securely fastened by means of seal-
ing wax into the cover b^ so that the platinum disks c^ c
always hold their relative positions.

For the better conducting solutions, a vessel of the form
shown in Fig. 15 is very suitable. It consists of the glass

PlO. 15

vessel a^ each arm of which is provided with a hard-rubber
cap b bearing the curved platinum electrode c,

67. PlatlnlzLngr the Electrodes. — The electrodes in
either vessel should be coated with a good layer of platinum
black, which is best obtained by introducing the clean
platinum electrodes into a 3-per-cent. solution of platinum
chloride containing -^ per cent, of lead acetate, and passing
the current from four Daniell cells for 5 or 10 minutes and
then reversing the current and passing it for an equal length
of time in the reverse direction. The electrodes must be
thoroughly washed before they are ready for use.

o2 ALKALIES AND HYDROCHLORIC ACID § :\0

68. Determination of tlie Conduetivlty of Solutions.

In determining the conductivity of solutions, use is made of
the apparatus described in Art. 63, except that on account
of the polarization (see Art. 83) by the passage of the cur-
rent, it is not possible to use a direct current. Instead of
the direct current, it is necessary to have a current that
flows at one instant in one direction, and the next instant in
the opposite direction, for by this means polarization can be
largely avoided. The alternation of the current can be pro-
duced by means of an induction coil that is introduced
between the battery By Fig. 12, and the Wheatstone bridge.
The difficulty then arises that the galvanometer cannot be
used, for the rapidly alternating current would simply cause
the needle of the galvanometer to tremble. Therefore, in
place of the galvanometer C, a telephone is used, which
gives a buzzing sound as long as a current flows through it,
and so shows when the branches of the bridge are equal.

Pig. 16

This arrangement is shown in Fig. 16. The battery B
furnishes the current to the induction coil /, where it is
made to alternate rapidly. The conductivity vessel is repre-
sented by r, and a known resistance by R. R and c make

§ 30 ALKALIES AND HYDROCHLORIC ACID 53

up two sides of the Wheatstone bridge and the wire e a b d,
which is stretched over a graduated scale, and has a sliding
contact/, makes up the other two sides. In making a deter-
mination, the solution is placed in the conductivity vessel r,
which vessel is put in a constant-temperature bath. The
resistance R is selected so as to be nearly equal to the
unknown resistance (or conductivity). (If the resistance of
the solution in c is totally unknown, a preliminary determi-
nation will show the approximate value of c, when R can be
suitably selected.) The induction coil /is then started and
the contact/" slid until the noise ceases in the telephone T.

By then reading the length of a and b on the scale, the ratios

d c

T = ^ are known. That is, we know a^ b and R^ and since

a 1

R-j^zc^ r is easily calculated and the conductivity equals -,

To get the specific conductivity, which is the conductivity of
a cube with 1 centimeter edge of the solution, it is necessary
to know the surface measurements of the electrodes and
their distance apart. This is not easy, however, so that use
is generally made of what is called the resistance capacity of
the vessel.

69. Resistance Capacity. — In order to determine the
resistance capacity, use must be made of some compound
that can be obtained in a pure state, of which a solution of
definite strength can be prepared and whose specific con-
ductivity is already known. In order, however, to obtain
accurate results, the resistance capacity of the vessel in
which the determination is made has to be ascertained.

Calling the specific conductivity of a certain solution /, that
of the same solution in the vessel used Z, and the desired
resistance capacity of the vessel K, we obtain the formula

In all further determinations with the same vessel, the
value K can be used, as it represents a constant so long as
the electrodes keep their relative positions. From this it
follows that having determined the conductivity of any

64 ALKALIES AND HYDROCHLORIC ACID § 30

other solution in the. same vessel, the specific conductivity
of any such solution may be obtained by the formula

l=KL,

70. Solutions for Resistance Capacity* — The follow-
ing solutions are suitable for use in determining the resistance
capacity of a vessel :

Sulphuric-acid solution, 30 per cent. H^SO^^ has a specific
gravity of 1.223 at 18® C. Ordinary chemically pure sul-
phuric acid is suitable for making the solution. The specific
conductivity at 18° C. is /= .7398. A ± error of .005 in the
specific gravity determination causes a ± error of .0004 in
the conductivity value.

A magnesium-sulphate solution has a specific gravity of
1.19. at 18° C. Commercial chemically pure magnesium sul-
phate is good enough for use. The specific conductivity at
18° C. is /= .04922. An error of .003 in the specific gravity
corresponds to an error of .00001 in the specific conductivity.

Other solutions are sometimes used, but these will usually
meet the needs of the worker in the electro-alkali industry.

71. Quantity of Electricity. — This is measured by
determining the amount of silver deposited by the current;
or, since there is a direct relation (see Art. 80) between the
amount of silver and any other metal that may be separated,
copper and, sometimes, hydrogen are separated instead of
the silver. A suitable arrangement for carrying out this
measurement consists of a copper plate or wire gauze that
can be accurately weighed and two other copper plates. In
measuring the quantity of electricity, the weighed plate is
hung, between the other two copper plates, in a solution of
15 grams of copper sulphate, 5 grams of sulphuric acid, and
5 grams of alcohol in 100 cubic centimeters of water. When
the current passes, the copper is dissolved from the outside
plates and deposited on the weighed one, so by weighing the
middle plate at the end of the process, the amount of current
that has passed can be readily calculated. Each coulomb
deposits. 329 milligram of copper and, therefore, the total
weight, in milligrams, of copper deposited divided by .329

§ 30 ALKALIES AND HYDROCHLORIC ACID 55

gives at once the number of coulombs of electricity that has
passed through the voltmeter. Another very popular style
of apparatus for this purpose consists in passing the current
through a solution of sulphuric acid, using platinum elec-
trodes, and measuring the gas evolved.

The same arrangement can be used for measuring
amperes; for, noting the time required to deposit on the
plate, we have all the informatio necessary for our calcula-
tion. For the number of coulombs divided by the number
of seconds required for them to pass gives the number of
amperes. For example, if the voltmeter shows 40 coulombs
in 40 seconds, then we have 1 ampere

12t Ajnmeters. — Although the preceding arrangement

is the most exact for the measurement of the quantity of the

current of electricity,

there are instruments,

known as ammeters,

which have a sufficient

degree of accuracy for

most technical work

and, on account of their

great convenience in

handling, are very

largely used. The in- ,

struments have been

given in a great number

. , r ... "o- 1'

of forms, but probably

the most convenient and accurate is that shown in Fig. 17,
known as the Weston ammeter.

The Weston instrument depends for its operation upon
the fact that if a coil, free to move, is pivoted in a mag-
netic field, it will swing round its axis when a current is
passed through it. In these instruments a rectangular coil
is delicately pivoted between the poles of a permanent mag-
net, and when a current flows through the coil, it is deflected,
carrying with it a pointer that swings over the scale shown
in the figure. The movements of the coil are counterbalanced

66 ALKALIES AND HYDROCHLORIC ACID g 30

by small spiral springs; the greater the current, the greater
is the deflection of the coil. The ammeter is inserted in the
circuit so that all the current will pass through.

73. Electromotive Force. — The electromotive force is
measured most exactly by using a standard cell of known
electromotive force and comparing the unknown electromo-
tive force with it. The best known standard element is
Clark's, which consists of a rod of zinc in a saturated solu-
tion of zinc and mercury sulphates, and has mercury for the
other pole. Such an element, when carefully made, has an
electromotive force of 1.4336 volts at 15° C. Thiscell varies
considerably with the temperature on account of the vary-
ing solubility of the zinc sulphate with varying temperature.
The high temperature coefficient is a decided disadvantage,
so that the Weston cell, which has a comparatively small
temperature coefficient, is becoming popular. It consists of
a cadmium amalgam in a saturated solution of cadmium
and mercury sulphates, with mercury for the other pole.

A very suitable
form of the Clark ele-
ment, and one that
can be conveniently
made in any labora-
tory, is shown in
Fig. 18. It consists of
a small glass cylinder a
set in a wooden block b,
and containing mer
cury c in the bottom
then a layer of mercu
rious sulphate d, cov
ered with a mixture ol
zinc-sulphate crystals e
and saturated zinc-sul-
phate solution f. A
cork g is then soaked
P'o- w in melted paraffin.

§ 30 ALKALIES AND HYDROCHLORIC ACID 57

and a zinc stick // and a glass rod /, in which a platinum
wire is fused, are fastened into the cork and the whole
inserted in the cylinder. A layer of wax is then placed
over the stopper, and wires lead from the platinum wire
and the zinc stick to the binding screws k^ k.

74, Measurement of Electromotive Force. — It is not

advisable to compare a number of cells direct with a standard
element, for it taxes the capacity of the element too much.
We can check up the capacity of a constant element against
the standard, and then use it for comparison. The deter-
mination depends on the fact that if we close a constant cell
with a resistance, the fall of potential will be uniform over
the whole length of the resistance. Furthermore, if a cell
is connected with another cell of equal but opposed electro-
motive force no current will flow.

The operation consists in closing the cell E, Fig. 19, with
a resistance a b. The fall of potential is uniform then for

Fig. la

each portion of a b. The wires connecting E with a and b
are so large that they have practically no resistance com-
pared with a b. From a a wire leads through the galvanom-
eter G and the unknown cell x to the slide contact c. At
intervals, to check the constancy of E, the standard cellis
introduced at .r, and c is moved until no current flows.
Then a r represents the fraction of the electromotive force
of E that is equal to the electromotive force Y of the stand-
ard cell. The standard cell is then replaced by the one to
be measured, and the point r, at which no current flows, is
again established. Calling this resistance ac\ then the

58 ALKALIES AND HYPROCHLORIC ACID § 30

electromotive force of the cell being measured is equal

ac'
to Y — . If a Clark standard cell is being used, F =1.4336
ac

ac'
and 1.4336 — = unknown electromotive force.
ac

A small storage battery is a very suitable cell for E^ and

the distance a c need only be determined twice a day.

75. The Voltmeter. — For a great many purposes an
instrument called a voltmeter is sufficiently accurate and
much more convenient for measuring electromotive forces
than the method just described. It is really an ammeter
having a high resistance and provided with a scale calibrated
to read volts instead of amperes. If we call the current r,
the electromotive force ^, and the resistance /?, then e =^ cR
(see Art. 77). Then, if the resistance of the instrument is
infinitely large compared to the resistance of the rest of the
current, the instrument having been calibrated to read
volts can be used to read direct. A voltmeter is connected
across the circuit, so that the entire current does not flow
through it.

76 Shunt Circuit. — When a wire leads continuously
from one side of a battery, or other source of current, to

the other side, it is called a
circuit. If, however, two
points of the circuit are con-
ffl nected by a wire, it is called
a shunt circuit. For exam-
ple, in Fig. 20 the wire act
forms a circuit from the
battery E. When a wire is
brought across from ^j: to ^, a shunt circuit, or shunt, is formed.
If the wire ab has a small resistance compared with ac b^
then the current will mostly pass across ab^ and in the
reverse case the opposite is true. If they are of equal
resistance, the current will be equally divided. If it is
desired to obtain the difference of potential between the

§ 30 ALKALIES AND HYDROCHLORIC ACID 59

points a and b^ a high-resistance voltmeter is inserted in the
shunt a b and the difference of potential is read direct.

77. Oliin's Iaw. — The relation existing between the
current, the electromotive force, and the resistance of a
system is known as Ohm's law. It is that the current is
directly proportional to the electromotive force and inversely
proportional to the resistance,

electromotive force

current =

resistance

78. Electric Conductors. — When an electric current
passes through a wire, the wire may become hot or suffer
other physical changes, but it remains essentially the same
as before. On the other hand, if the current passes through
a solution, it decomposes the dissolved substance and its
products collect at the points where the current enters and
leaves the solution. This leads to a division of electric con-
ductors into two classes. All electric conductors that are
not decomposed by the electricity passing through them
are called conductors of the first class; all conductors that
are decomposed by the electricity passing through them are
called conductors of the second class, or electrolytic con-
ductors.

There is an indefinite number of conductors of the second
class, most of which may, however, be comprehended in
the general title of solutions. Comparatively few pure
substances other than the metals conduct electrolytically.
Such substances as hydrochloric, nitric, and sulphuric acids,
which in water solution are good conductors, do not conduct
at all when in the pure, dry condition. By the pure, dry
state is meant hydrochloric-acid gas condensed to a liquid
and mixed with no other substance ; the same is meant for
nitric acid and sulphuric acid. Water is also a very poor
conductor. Fused salts, however, conduct quite well and
some few, as lead and silver chlorides, conduct somewhat in
the solid condition when not too far from their melting
point.

60 ALKALIES AND HYDROCHLORIC ACID § 30

ELECTROLYSIS

79. As stated above, sulphuric acid, although a non-con-
ductor when pure and dry, when dissolved in water is a good
conductor and the solution is an electrolyte. Solutions in
general that conduct are called electrolytes, although the
term is frequently applied to the dissolved substance. For
instance, in the above case it is customary to speak of sul-
phuric acid as an electrolyte, meaning that its water solu-
tion is a good conductor.

The current enters and leaves the solution by wires, and
these, where they dip into the solution, are called elec-
trodes. When a current is passed through a sulphuric-acid
solution, oxygen separates at one electrode and hydrogen at
the other. The electrode at which oxygen, or, in general,
the acid radical, separates is called the positive electrode^ or
anode; and the one at which hydrogen, or, in general, the
metallic radical separates, is the negative electrode^ or
cathode. Since by the passing of an electric current through
an electrolyte, matter separates out at the electrodes, the
electrolyte must be decomposed and matter must be carried
with the current, for the concentration about the electrodes
soon differs from the rest of the solution.

The matter that travels with the current is called ions.
The ions that travel towards the anode are called anions^
and those that travel towards the cathode are called cations.
The ions are perfectly definite substances, but frequently
they are not the substances that separate at the electrodes,
for at the instant they are set free they may react with the
solvent to form new substances ; for example, the ions from
sulphuric acid are hydrogen and SO ^^ the hydrogen separates
as such, but the SO ^ breaks down and gives oxygen and sul-
phuric acid once more. In the electrolysis of sodium sulphate
the ions are sodium and SO^^ but the sodium reacts with the
water to give hydrogen and sodium hydrate, and the SO^
acts as in the above case, giving oxygen and sulphuric acid.

80. Fai-aday^s Law. — When a certain am^ount of elec-
tricity passes through a solution of sulphuric acid, a definite

§ 30 ALKALIES AND HYDROCHLORIC ACID 61

amount of hydrogen is liberated ; and for the same amount
of current, the same amount of hydrogen is liberated inde-
pendent of the rapidity or slowness with which the current
acts, the concentration of the solution and the temperature.
Each gram of hydrogen liberated by an electric current cor-
responds to the passage of 96,540 coulombs; it makes no
difference what substance is electrolyzed to give hydrogen,
so long as only hydrogen is liberated at the cathode 1 gram
will be freed when 96,540 coulombs of electricity has passed.
If we, therefore, pass an electric current successively through
solutions of hydrochloric acid, sulphuric acid, phosphoric acid,
etc., exactly the same amount of hydrogen will be liberated.
What has been stated for hydrogen holds true for other
elements and combinations of elements. If in electrolyzing
a solution of sulphuric acid the hydrogen given off at the
cathode and the oxygen at the anode (having waited until
secondary reactions, which appear at the beginning of the
electrolysis have stopped) are measured, it is found that the
volume of the hydrogen is twice that of the oxygen. That
is, the gases are liberated in the proportions in which they
combine. Equivalent weights of the substanees are liberated.
Furthermore, if an electric current is passed successively
through solutions of sulphuric acid, copper sulphate, silver
nitrate, ferrous sulphate, and ferric sulphate, if the solutions
are suitably prepared to avoid secondary actions at the elec-
trodes, we will get, when 1 gram of hydrogen is liberated,
3L5 grams copper, 108 grams silver, 28 grams of iron from
the ferrous solution, and 18.7 grams iron from the ferric
solution. If the atomic weights of these elements are
noticed, it will be found that the above values are in each
case the atomic weight of the element expressed in grams
divided by its valence. This relation was first noticed by
Faraday and is known as Faraday's law. Briefly stated, it
is that chemically equivalent quantities of substances are
separated by the same amount of an electric current. For
every 96,540 coulombs of current that pass, if no side
reactions enter in, 1 gram equivalent each of the cation and
of the anion is obtained.

62 ALKALIES AND HYDROCHLORIC ACID § 30

81. Electrolytic Dissociation* — The way' in which the
current is carried in an electrolyte has long been a subject
for speculation. It is now possible, however, to account for
the quantitative phenomena of electrolysis by assuming that
the dissolved substance is dissociated before the passage of
the current. For example, when sodium chloride is dis-
solved in water, it is dissociated to a greater or less extent
into sodium ions and chlorine ions, each of which bears a
charge of electricity. That substances are so dissociated is
also made very probable by measurements of the boiling and
freezing points of solutions of electrolytes. Now, when the
current passes through the solution, for every 96,540 cou-
lombs of electricity passed, a gram equivalent of the cation
and of the anion separates out, gives off its charge at the
proper electrode, and becomes an ordinary substance again.

Substances are usually not entirely dissociated in solution,
but consist of a mixture of undissociated and dissociated
molecules. In water solution, most of the salts and the
stronger acids and bases are quite highly dissociated at mod-
erate dilution, and the dissociation ranges from this to zero
dissociation for non-conductors.

Since the electricity is carried by the ions, its conductivity
by a solution must depend on the number of free ions in
solution and the speed with which they move. An increase
in the concentration of a solution increases the number of
free ions and its conductivity, but this conductivity is not
proportional to the increase in concentration, for the more
concentrated a solution is, the less is it dissociated.

83. Ml^rratlon Velocity. — The speed with which the
ions move depends on the viscosity of the solvent and the
individual kind of ion. The speed with which some ions
travel at 18° C. in water solution, with a difference of poten-
tial between the electrodes of 1 volt, is given in Table II.

It will be seen that the velocity with which ions move
through water varies considerably, hydrogen and hydroxyl
far exceeding all others. Hydrogen moves about five
times as rapidly as chlorine, so that in the electrolysis of

§ 30 ALKALIES AND HYDROCHLORIC ACID

hydrochloric acid there is a tendency for the concentration

of the acid to rapidly decrease at the anode and increase

at the cathode.

TABLE n

Cations

Centimeters
per Hour

Anions

Centimeters
per Hour

H
K

Na
Ag

lo.So

2.05
1.98
1.26
1.66

NO,

C,H,0,

5.60
2.12
2.19
1. 91
1.04

83. Polarization. — When a suitable electric current is
passed between copper electrodes through a zinc-sulphate
solution, the copper dissolves from the anode to form copper
sulphate, and the zinc is deposited on the cathode. If after
the current has been passing for some time, the source of
current is cut out and the copper plates are connected by a
wire, a current will flow in the opposite direction from the
first one. This phenomena is known as polarization. If
the electromotive force of the above cell is measured, it will
be found to be about 1.1 volts. That is, a Daniell cell has
been formed, and the condition very soon after the direct
current begins to pass is the same as if a current were
running against a Daniell cell. Therefore, it will not be
possible to keep up the passage of electricity through a cell
of this kind unless the original current has an electromotive
force greater than the electromotive force of polarization.

The passage of a current through copper electrodes in a
solution of copper sulphate simply dissolves copper from the
anode and deposits it on the cathode, so that in this case
there is no polarization and a current of the smallest electro-
motive force will flow continuously.

Almost all cases of electrolysis give polarization, and the
passage of the current can only be continued when the elec-
tromotive force of the source of the current is greater than

64 ALKALIES AND HYDROCHLORIC ACID § 30

the electromotive force of polarization of the solution. This
electromotive force of polarization can be measured directly
or it can be calculated from the heat of the reaction that
would cause the polarization.

84r, Calculation of the Electromotive Force of
Polarization From the Heat of Reaction. — When a
metal reacts in an electric cell, there is a certain amount of
energy set free that may be evolved as heat or as electric
energy, as circumstances may favor the one or the other.
If, therefore, the heat of the chemical reaction is known,
the electromotive force of the cell can be approximately cal-
culated. This will, however, not be the exact value for the
cell, for a temperature coefficient, which varies with the
kind of cell, also enters into the calculation. Since the elec-
tromotive force of polarization is only the current tendency
set up by the separated product, it can also be calculated in
the same way as the direct electromotive force.

If we call the electromotive force of polarization e and
represent the valence of the ion by «, then, when 1 gram ion
has separated out, or if we have the gram ion formed
from the electrode and going into solution, we have the
electrical energy // r 96,540 volt coulombs. This, in calories,
is ;/^ 90,540 X .24 = «r 23,170 calories. If we represent the

heat energy by Q, then Q = « ^23,170 and e = r^ volts.

We can calculate from this very nearly the minimum elec-
tromotive force necessary to electrolyze a solution, assuming
that no secondary reactions enter in. For example, if a
solution of hydrochloric acid is electrolyzed, hydrogen sepa-
rates at one pole and chlorine at the other. These, from
their tendency to combine, will give an electromotive force
opposed to the decomposing current, which can be calculated
by the above formula. The heat of formation of a gram
molecule of hydrochloric acid in dilute solution is 39,300 cal-

39 300
ories and the valence of hydrogen is 1 ; therefore e = ' *

= 1.69 volts and it will require a current of at least that

§ 30 ALKALIES AND HYDROCHLORIC ACID 65

electromotive force to pass continuously through such a
solution.

85, Summary. — We may briefly summarize our ideas
about the electrolysis of a solution in the following laws :

1. Every electrolyte is, by the passage of the current,
decomposed into two parts — the cation and the anion.
These are in certain cases the positive and negative ele-
ments of the compounds, as in sodium chloride where the
ions are sodium and chlorine; in other cases they are com-
binations of elements, as in potassium ferrocyanide, where the
cation is potassium and the anion is the ferrocyanide radical.

2. The metal of a compound usually separates at the
cathode, but in certain cases, as in ferrocyanides, one metal
goes to the anode.

3. Water solutions of salts of the metals that decompose
water naturally do not give the metal at the cathode, for as
soon as the metal is separated it decomposes the water and
forms a hydrate. Very strong solutions of the hydrates may
be exceptions to this; also when a mercury cathode is used
the metal dissolves in the electrode and is protected from
decomposition.

4. The liberated ion appears only at the surface of the
electrode.

5. There is a certain minimum electromotive force
required for the electrolysis of a solution, which is deter-
mined by the heat of reaction of the liberated ions. If less
than this minimum electromotive force is supplied, the
current will pass until enough of the ions are liberated to
set up the electromotive force of polarization, when the cur-
rent will stop. In the case of the electrolysis of a solution
between electrodes of the same metal as the positive ion,
there will be no polarization and the weaker current will
flow continuously.

6. The chemical work done is proportional to the mini-
mum electromotive force of polarization, and if a greater
electromotive force than the minimum is required, it will

66 ALKALIES AND HYDROCHLORIC ACID § 30

not appear as chemical work in separating more ions, but as
heat energy.

7. Various secondary reactions may take place as: (a)
The decomposition of one or both of the ions (usually the
negative one, however). For example, SO^ may decompose
into SO^ and O. (d) The ions may react on the electrodes,
as in the electrolysis of dilute sulphuric acid between zinc
electrodes, in which case the SO^ acts on the anode, giving
ZnSO^y and only hydrogen is set free, {c) Abnormal ions
may be liberated, as the frequent formation of ozone (?„ the
deposition of a black porous deposit of copper and the
deposition of lead or manganese dioxide on the anode.

EliECTROIiYTIC PREPARATION OF ALKAIil

ANT> CHIiORIKE

INTRODUCTORY

86. Historical. — The fact that solutions are decom-
posed by the electric current has been known since the
beginning of the 19th century, and a process was patented
for the electrolysis of salt solutions during the first half of
that century. It was not until the dynamo was perfected,
however, that the commercial electrolysis of salt solutions
could even be considered. About 1880 an interest in the
subject began to be shown by applications being made for *
patents; but even in 1888 many leading men in the alkali
industry considered the electrolysis of salt in a commercial
way impractical. At the present time, however, there are
several processes that may be considered commercially
successful for the making of alkali and bleach from salt by
electrolysis, and more than half of all the chlorate of the
world is made by this method.

87. Electrolysis of Salt. — This involves first the
separating of the ions — sodium on the cathode and chlorine
on the anode. Then, if we are electrolyzing fused sodium
chloride, the chlorine is evolved and collected, and the

§ 30 ALKALIES AND HYDROCHLORIC ACID 67

sodium separates as metal ; if the temperature is kept suit-
ably high, it can be drawn off and cast into bars. This
process might be used for the preparation of metallic
sodium, but it is possible to produce the metal more
cheaply and easily by the electrolysis of the fused hydrate.
If a solution of salt is used for electrolysis, the chlorine will
be evolved as before, but the sodium acts on the water as
soon as set free and forms sodium hydrate and hydrogen.
As soon as formed, the caustic-soda solution begins to
conduct a portion of the current, and to be decomposed,
liberating oxygen at the anode and wasting the current.
There is also a possibility that a portion of the chlorine
will get mixed with the caustic liquor, and so form sodium
hypochlorite, which may, in turn, be converted into sodium
chlorate or be reduced by the hydrogen to sodium chloride.
These various processes may be represented by the equations

"iNaCl = %Na + C/,
%Na + %Hfi = 2NaOI/+ H^
%NaOH-^ Cl^ = NaCl-\- NaClO + Hfi
ZNaClO = %NaCl + NaClO^
NaClO-^H^ = NaCl + H^O

In addition to the loss of alkali and chlorine by its reversion
to salt, we must remember that, as was pointed out with the
sodium hydroxide, these substances all conduct and waste
current.

88. Conditions Favorinfir Electrolysis. — The ideal
conditions towards which we must aim in selecting a proc-
ess for the electrolysis of salt, for the formation of sodium
hydrate and chlorine, may be summarized as follows :

1. The process must work at as low a voltage as possible,
in order to give the maximum decomposition per electrical
horsepower.

2. The combination of the caustic soda and chlorine to
form sodium hypochlorite must be avoided, in order to pre-
vent a loss of current and to avoid great wear and tear on
the electrodes. The accumulation of the sodium hypo-
chlorite also prevents the continuous use of the electrolyte.

68 ALKALIES AND HYDROCHLORIC ACID § 30

3. The products of the electrolysis must not be allowed
to accumulate in the decomposition cell.

4. Strong and pure solutions of sodium hydrate must
be obtained, in order to avoid the expense of concentrating
the solutions and that the product may be salable.

5. The apparatus must be simple and need but little
attention and repairs.

89. Electrodes. — The cathodes in the electrolysis of
salt solutions cause very little trouble, as it is compara-
tively easy to find materials that are resistant to the action
of caustic soda*. With the anode it is, however, much dif-
ferent, for here is set free the very active chlorine, and, by
secondary actions, the still more active oxygen and oxides
of chlorine. The obtaining of anodes that would be suffi-
ciently resistant, and at the same time not too expensive,
was in the early days of this work one of the most difficult
problems to solve.

The two conditions that a successful electrode must fulfil
afe that it shall be a good conductor and at the same time
resistant towards the products of electrolysis. The only sub-
stances that satisfactorily meet these conditions are carbon
and the platinum metals, with their alloys. Carbon, in the
form of coke, is not badly acted on by chlorine, but oxygen
and the oxides of chlorine act on it considerably and cause
it to disintegrate. The overcoming of this difficulty was at
one time almost despaired of, and recourse was had to
making the electrodes as cheaply as possible from slabs of
gas coke and frequently renewing them. At the present
time, however, carbon electrodes are made by mixing finely
ground coke with tar and some suitable metal or metallic
oxide, pressing it into shape and heating it to drive off the
more volatile substances. The electrodes are then subjected
to the highest temperature of the electric furnace. By
this means, carbides of the metal are formed, which are
immediately decomposed with liberation of the metal, and
the carbon is left behind in a fine graphite form. Carbon
electrodes made by this, or a similar method, are now very

§ 30 ALKALIES AND HYDROCHLORIC ACID 69

generally used in the production of caustic soda and chlorine
by electrolysis.

The other possibility for anodes is an alloy of 90 per cent,
of platinum and 10 per cent, of iridium, which is far more
resistant towards the products of electrolysis than platinum
alone. These electrodes are expensive, however, and are
not so much used in the preparation of chlorine and caustic
soda as the carbon electrodes. On the other hand, in the
preparation of chlorates the platinum iridium alloy is almost
exclusively used, as the use of carbon is practically out of
the question on account of the oxidizing substances formed
in large amounts.

FUSED ELECTROIiYTE

90, The use of fused salt as an electrolyte offers certain
difficulties that do not occur with the solution, and inventors
have largely turned their attention to the perfecting of those
processes that use solutions of salt in water. Three of the
processes using fused salt as an electrolyte which have been
patented deserve mention; they are Vautin'Sy Hulin*s, and
Acker*s, Of these processes, Vautin's proved impractical and
has apparently been abandoned, but the other two processes
are in apparently successful operation.

inJLTN»S PROCESS

91. Hulln's process consists in the electrolysis of a
fused mixture of sodium and lead chlorides, using a lead
cathode. One difficulty that is experienced ordinarily in the
electrolysis of fused salt is, that both the sodium and chlo-
rine rise to the top of the material and it is very hard to pre-
vent loss by their reuniting. In this method, however, the
lead cathode is fused, and at the bottom of the electrolyte, so
that the chlorine is evolved and carried away from the top of
the apparatus and the sodium remains as an alloy with the
lead in the bottom. Vautin employed a similar arrangement,
but attempted to electrolyze sodium chloride alone; this led
to the formation of a crust of the lead-sodium alloy on the

70 ALKALIES AND HYDROCHLORIC ACID § 30

surface of the cathode, with a subsequent high electromotive
force and loss of sodium. Hulin avoids this difficulty by using
an electrolyte of a mixture of sodium and lead chlorides, so
that lead is continuously deposited with the sodium and an
alloy of the proper composition built up. By this method,
the mixture of chlorides must continuously become poorer in
lead chloride, unless more of the substance is continuously
added. This addition of lead chloride is best made, or rather
the lead for the cathode is best supplied (for it consists in a
simple transfer of lead from the anode to the cathode), by
employing two anodes, one of carbon, the other of lead.

By allowing any desired fraction of the total current to
pass through the lead anode, as much of it as is needed is dis-
solved in the electrolyte. It is found in practice that the best
results are obtained by allowing 12 per cent, of the total cur-
rent to pass through the lead anode and the remainder through
the carbon anode. The electrolysis takes place in cast-iron
crucibles, which are surrounded by bad heat-conducting
material and lined with an insulator. The heat of formation
of salt from sodium and chlorine is 97,600 calories and there-
fore, according to the formula le = 90^^170) ^^^ electromo-
tive force theoretically necessary to decompose fused sodium
chloride is about 4.2 volts, for this value is calculated using
the heat of formation of solid sodium chloride ; that for the
fused chloride will be less by the heat of fusion, and its elec-
tromotive force of polarization will also be less. In practice,
each crucible employs a current density of 700 amperes per
square foot of electrode surface and 7 volts electromotive
force. By the use of such high current density it is possible
to get a large amount of decomposition of the electrolyte per
unit of electrode surface, and thus to employ a small plant.

The yield per electrical horsepower hour is 81 grams of chlo-
rine and 54 grams of sodium. The chlorine is converted into
bleaching powder by the usual method. The lead alloy, which
contains from 23 to 25 per cent, of sodium, may be sold directly
for many uses where metallic sodium is required. It is usually,
however, treated with water, and by suitable working, a strong

§ 30 ALKALIES AND HYDROCHLORIC ACID 71

solution of caustic soda of a high degree of purity is obtained.
This caustic requires very little fuel for its evaporation, and
for this reason is much better than the more dilute caustic
obtained by many processes. The lead is left by this oper-
ation as a spongy mass, and together with considerable lead
peroxide that is also formed, it makes a valuable by-product.
This process was considered so promising in 1899, that a
company was formed with a capital of over 1500,000, and
works, which are still in successful operation, were erected at
Clavaux, France, for carrying out this method.

THZ ACKEB PROCESS

93. The Acker electrolytic proceee, which is at
present in successful operation at Niagara Falls, differs from
the above in that it
uses fused lead as
the cathode and
continuously re-
moves the sodium
from the sodium-
lead alloy, so that
the lead can be used
continuously. The
apparatus for car-
rying out this proc-
ess is shown in
Fig. 31. It consists
of an iron base a
embedded in brick-
work b, which rest
on brick pillars c,
or it rests on the
ground and has

places excavated ^"^- "

for the parts projecting below the surface. The upper part
consists of slabs d oi acid-resisting slate or is made of fire-
clay. These slabs are carefully luted into the iron shoulders,

72 ALKALIES AND HYDROCHLORIC ACID § 30

as shown, by using fireclay. Through the top cover project
the graphite anodes ^, e^ e^ while at / is provided a charging
hole for fresh salt. At g is molten salt and at h an alloy of
molten lead and sodium. At i is a pipe for conducting away
the chlorine. At y is a pipe for blowing in steam ; k serves for
conducting away the hydrogen; /conducts the fused caustic
soda 0 to the shipping tin ;;/. The extension/ serves for draw-
ing away the fused contents of the cell, when it is necessary
to empty it for repairs, and q shows the cathode connection.
At s is the iron plate that serves to separate the molten lead,
which is the cathode proper, from the alloy below. The top
is covered with a non-conducting material /, as asbestos wool.
To start the operation, the interior of the cell is heated
by hydrogen flames until it is thoroughly hot ; then melted
lead and melted salt are run in, the covers and electrodes
put in place, and the current started. The chlorine is
given off at the anodes, rises to the surface, and is con-
ducted away through the pipes /. The sodium separates on
the surface of the fused lead, which acts as the cathode, and
alloys with it. Meanwhile, superheated steam is blown in
through/ and causes the lead to rise in w^ overflow, and cir-
culate as shown by the arrows. As soon as the cell is in
working order, the sodium alloy is decomposed in zv by the
steam and the fused caustic soda rises to the surface of the
lead in o and runs off through / into the shipping can ///;
/ contains a plunger valve so that the flow of caustic can be
stopped if desired. The lead flows in the direction of the
arrows, displaces the sodium alloy just formed, and so forms
a system of circulation. The hydrogen, which is formed
by the action of the steam on the sodium in the alloy,
escapes through k and can be collected and burned over the
salt, in the form of an oxyhydrogen flame, to keep up
the temperature of the cell. Since the cell is well insulated,
the heat from the steam and the heat of the reaction

Na + HP = NaOH^r H

nearly suffice to keep up the temperature of the cell to the
proper point.

§ 30 ALKALIES AND HYDROCHLORIC ACID 73

I>I880IiVED ELECTROLYTE

93. The so-called wet processes, or those in which the
sodium chloride is in solution as an electrolyte, comprise
the most important methods for obtaining the products of
electrolysis. A serious difficulty is encountered in working
these processes, by the materials formed at the electrodes
tending to mix and form compounds that are not wanted.
The methods used for keeping separate the products formed
about the electrodes may be divided into three general
classes: (1) By a difference in the density of the liquids;
{2>) by diaphragms; (3) by using a mercury cathode

DIFFERENCE IN DENSITY

94. The processes depending on the difference in the
specific gravity of the sodium hydrate formed and the
rest of the solution, to keep the products of the reaction
separate, place the anode at the top of the decomposition
vessel, so that the chlorine is set free without traversing
more than a small portion of the liquid. On the other hand,
the cathode is placed at the bottom of the cell, and the
caustic solution being heavy stays at the bottom and can be
drawn off. Theoretically, this is a good arrangement, but
practically it is almost impossible to prevent the diffusion
and mixing of the chlorine and caustic soda. This difficulty
is also increased by the hydrogen, which is set free at the
cathode, rising through the electrolyte and mixing it. The
Richardson-and' Holland process avoids the difficulty with
the hydrogen by using a copper cathode covered with a
coating of copper oxide. The copper oxide oxidizes the
hydrogen as rapidly as it is formed. When necessary, the
electrodes are removed and the copper oxide is regenerated
by heating in the air. By this method a fairly good sepa-
ration of the caustic soda and the chlorine can be main-
tained; this process was tried on a manufacturing ,scale,
but it has been abandoned.

74 ALKALIES AND HYDROCHLORIC ACID g 30

I USING DIAPHItAOMS

95. The use of a diaphragm is a favorite device for
keeping the solutions around tke cathode and anode sepa-
rate, but it is very difficult to find a diaphragm that will
meet all the requirements. A diaphragm to be satisfactory
must resist the action of the contents of the bath, must keep
the anode liquor well separated from that of the cathode,
and must not offer great resistance to the passage of the
current. A large number of diaphragms have been pro-
posed, but as none of them has been very satisfactory, it
will be sufficient to consider two in connection with the
most important processes of this type. Of the various forms
of apparatus using diaphragms for electrolysis of a salt solu-
tion, Greenwood's is among the most satisfactory.

96. GreenMTood Process. — Greenwood's apparatus con-
sists of a circular iron cell a. Fig. 33, lined with copper,

which also serves as the
cathode. On the bottom
of the cell is placed a slate
slab d, upon which rests
the diaphragm c, which con-
sists of a series of V-shaped
circular troughs of glass,
or porcelain, and packed
together by asbestos. The
anode is a carbon piece
with a core of type metal
and stands inside the cir-
cular diaphragm. When
in operation, a series of
these cells are placed step-
wise, as shown. The brine
; enters continuously at </
\ and flows into the next
cell through y"; the chlorine
^'°' ** escapes through e and the

caustic goes to the next cell with the brine. Theoretically,

§ 30 ALKALIES AND HYDROCHLORIC ACID 75

aboat 2 volts are required for electrolysis, but in practice
i.i volts and a current density of 100 or 110 amperes per
square meter are used. The electrolyzed brine contains
about 10.76 per cent, of salt and 2.3 per cent, of caustic
soda. This must be evaporated down and the salt 6shed
out. The process was in experimental operation in Eng-
land as late as 189fl, but it is very doubtful if any process
that gives such a weak caustic, mixed with so much salt,
can prove a success, for the expense of concentrating the
solution and fishing out the salt is too great.

97, Le Sueur Process. — This is a combination of
the density and diaphragm methods of separation, for while

it uses a diaphragm, the electrodes are so placed that the
gravity separation will be as effective as possible. The
electrolyzing vessel a, Fig. 23, consists of J-inch boiler steel
and is about 9 feet long, 5 feet wide, and IJ feet deep. The
anode compartment is made by building up red bricks d in
Portland cement somewhat higher than the electrolysis cell
and covering it over with spruce planks c. Carbon has been
discarded as an anode substance in favor of the 10-per-cent.
iridium-platinum alloy already referred to.

The anodes are made according to a method devised by
Le Sueur, which consists in rolling 4-inch pieces of the
platinum-iridium wire very thin, except at one end; the
unrolled ends are then bunched together and fastened in a
glass tube, so that they just extend into the interior and
the flat ends spread out. When the anodes are in place
through the spruce cover to the anode compartment,

76 ALKALIES AND HYDROCHLORIC ACID § 30

connection is made with the main conductor by means of a
drop of mercury in each glass tube, an iron wire reaching
to Ihe top of each tube. These electrodes cost about
73 cents each and sufficient to make 200 tons bleach per
month cost about 15,000.

The anode compartment is separated from the cathode
compartment by an asbestos diaphragm supported on a
wire gauze, which at the same time serves as the cathode.
By thus bringing the diaphragm close to the cathode, the
resistance of the cell is diminished ; and by making use of
the gravity system, the caustic soda is kept quite well
separated from the chlorine. It is nevertheless impossible
to prevent some diffusion and the formation of sodium hypo-
chlorite, which not only causes loss of current but also acts
on the electrodes. This is avoided in the anode compartment
by keeping the solution slightly acid with hydrochloric acid,
which decomposes the hypochlorite and gives chlorine. . The
sodium hypochlorite that collects in the cathode compart-
ment is converted into sodium chlorate and recovered. The
diaphragm and cathode are arranged as shown and are also
sloped to one end of the cell, so that the hydrogen passes
to the higher parts and then out of the cell. The diaphragms
last on an average 7 weeks, but have lasted as long as
24 consecutive weeks. The anodes and the cell itself are
practically indestructible. Instead of the theoretical 2 volts,
the process uses 6^ volts and 1,000 amperes per cell. A
solution containing from 10 to 15 per cent, of sodium
hydrate can be separated by this process, but it also con-
tains considerable salt. This liquor is concentrated under
diminished pressure, the salt separated by centrifugal
machines, and the evaporation completed in iron pots.

The efficiency of the process is about 87 per cent, of the
theoretical amount of chlorine and somewhat less of sodium
hydrate. The process is in successful operation on a com-
mercial scale at Berlin Falls, New Hampshire, where the caus-
tic is used in making wood pulp and the chlorine used to bleach
the pulp. It is very doubtful if any process depending on
either gravity or a diaphragm for the separation of the

§ 30 ALKALIES AND HYDROCHLORIC ACID 77

caustic soda can be successfully operated to make solid caus-
tic soda, for the expense of concentrating the necessarily
dilute solution and the separation of the salt is too great,
even if we leave out of consideration other objections.

98. Hargrreaves-and-Blrd Process. — This process can
best be classed under the head of diaphragm processes,
although strictly the diaphragm does not divide the cell.
The process is distinctive, in that the walls of the cell are
composed of the diaphragm and cathode. The diaphragm is
composed of a layer of paper or some other suitable material,
as a copper-wire gauze, covered with a layer of Portland
cement, which, in turn, is covered with a layer of asbestos.
This is impermeable to the salt solution, but allows the
sodium ion to pass. The cell is put together with a copper-
wire gauze, which serves as the cathode, on the outside, and
the whole is set into an enclosing jacket. The carbon
anodes are hung in the anode compartment and the brine to
be electrolyzed slowly flows in at the bottom of the cell and
passes out at the top, through the same pipes as the chlorine.
During electrolysis the sodium ions migrate to the top
cathode and are there, as rapidly as set free, converted into
caustic soda by blowing in steam; or into soda crystals, by
steam and carbon dioxide.

The diaphragm and cathodes are made 10 feet long and
5 feet high, and as one is on each side of the cell, it gives
100 square feet of cathode surface. A cell of this size
decomposes on an average 237 pounds of salt every 24 hours
and gives 365 pounds of 37-per-cent. bleach and 213 pounds
of soda ash by the use of 2,300 amperes and 3.9 volts per
cell. This represents an efficiency of about 97 per cent, of
the electrical energy used. The brine is best obtained
direct from the wells; in passing through the cell, 75 per
cent, of it is decomposed. The dilute brine can be returned
to the well to be resaturated. The chlorine can be converted
directly into bleach and the caustic is strong and pure.
When sodium carbonate is made, for the making of which
this process is well suited, the solution is so concentrated

78 ALKALIES AND HYDROCHLORIC ACID § 30

that the carbonate crystallizes out without concentration.
The sodium carbonate so made is very pure, averaging
when dehydrated 97.9 per cent, of NafiO^^ 1.53 per cent,
of NaCly and .53 per cent, of Na^SO^y etc. The sulphate is
probably due to sulphur dioxide in the furnace gases that
are used for carbonating.

The apparatus is simple and requires very little attention.
The only part that suffers great wear is the diaphragm, and
that is quite cheap. This process has been running satisfac-
torily in a small way for several years, but now a large
plant is being established in England.

PIUKJESSKS USING A MERCURY CATHODE

99. A large number of processes for the electrolysis of
salt have been proposed in which a mercury cathode is used.
These have the advantage that the sodium separates with
the mercury as an amalgam and can be converted into
hydrate outside of the cell. By this means a solution of
caustic soda of high concentration and practically free from
salt can be made. The process suffers from the disadvan-
tage that only dilute amalgams can be made, for otherwise
there is a loss of current, and therefore the mercury must
be frequently changed. There is also a chance of a large
loss of mercury, for when the sodium is acted on by the
water, mercury is mechanically carried away by the hydro-
gen ; also, considerable mercury is carried off in the form
of vapor, even at ordinary temperatures.

100. Castner-Kellner Process. — This is the most sat-
isfactory and successful process of this character ; in fact,
it has proved itself the most satisfactory of all processes for
the electrolytic decomposition of salt.

The cell is divided into three compartments, the center of
which contains the iron cathode a. Fig. 24, and serves for
the decomposition of the sodium amalgam. The two end
divisions serve as anode compartments and contain the carbon
anodes ^, b. One end of the cell rests on a knife edge r, and

g 30 ALKALIES AND HYDROCHLORIC ACID 79

the other is supported on the eccentric d, which revolves and
thus slowly raises and lowers the end of the cell. Brine fills
the two end compartments and is renewed, as necessary, by
fresh brine flowing in ; the exhausted brine goes to be resatu-
rated. A thin layer of mercury covers the bottom of the
apparatus, and is so regulated in amount that all of it prac-
tically flows alternately from the end compartments into the
middle, as the cell rocks.

Strictly speaking, the ends of the cell are not anode com-
partments, but are alternately complete cells in which the
salt is decomposed, the chlorine separating on the carbon
anode and passing off, the sodium dissolving in the mercury
cathode to form an amalgam. Then, as the cell tips, the

amalgam flows into the center compartment, where it forms
the anode of a primary battery, and the iron electrode here
becomes the cathode of this battery. This has the advan-
tage that the hydrogen, instead of coming from the surface
of the mercury and so carrying that metal with it, comes
from the iron cathode, and the sodiumsimply goes into solu-
tion from the mercury as caustic soda. This also has the
advantage that the current from this battery aids in the
electrolysis in the end cells. Owing to the frequent removal
of the sodium amalgam from the anode cell, it rarely con-
tains over .03 percent, of sodium, and as a consequence the
cell gives a high degree of efficiency, being from 88 to 90 per
cent, of the theoretical.

80 ALKALIES AND HYDROCHLORIC ACID § 30

Since no caustic soda is formed in the anode compartment
there is no formation of sodium hypochlorite, and therefore
the anodes have practically no wear. And since the elec-
trolyte contains no hypochlorite, it can be used continuously
by being conducted through a supply of salt so as to be
resatu rated. The resistance in the cell is very low, so that
a current of 4 volts and 550 amperes per cell will decompose
56 J pounds of salt every 24 hours and will yield 38|^ pounds
of caustic soda and 34^ pounds of chlorine. The caustic
solution can be made of almost any desired concentration
and is practically made about W per cent, sodium hydrate.
It can be concentrated by simple evaporation and yields
a caustic 99^ per cent. pure. The chlorine obtained is
from 95 to 97 per cent, pure, and for the rest contains a
small amount of hydrogen. The cells are very simple and
require but little attention, the work being almost auto-
matic. Repairs are seldom needed, but when necessary any
cell can be cut out from action without disturbing the work
of the remainder. This process has been working with
apparent success for several years in England, on the Con-
tinent, and in America. The English company has been
able to declare 8-per-cent. annual dividends on a capital of
over H million dollars.

101. Conclusions. — To sum up, then, we may conclude
about as follows :

1. In the fusion processes, fused salt is a good con-
ductor of electricity, and therefore very high current den-
sities can be used, which means that a larg^ output can be
obtained from a small plant. Concentrated solutions of
caustic soda, or, in the Acker process, even fused caustic
soda can be made. On the other hand, the wear and tear on
the cell, especially if heated from without, is very great and
the cost of keeping the material fused must be considered.
The hot chlorine is not so easy to handle as the cold chlorine
from the other processes.

2. The process using gravity for separating the products
has very few good points.

§ 30 ALKALIES AND HYDROCHLORIC ACID 81

3. In the diaphragm processes the cells are cheap and
the wear and tear on the cell is not great. They require
very little skilled labor. They suffer, on the other hand,
considerable loss of caustic soda and chlorine through their
recombination and by reduction at the cathode, and have
high resistance in the cell. This is, however, nearly pro-
portional to the power of the diaphragm to stop diffusion, so
the higher the resistance, the smaller is the loss of the prod-
ucts through mixing, and the reverse. They furnish a low
strength of caustic, the concentration and purification of
which is expensive.

The Hargreaves-and-Bird process cannot be included in
this general statement, as it is not strictly a diaphragm
process.

4. The mercury-cathode cell has very little loss through
the recombination of the products of the reaction. The cells
are quite free from wear and tear. A highly concentrated
caustic-soda solution can be made if desired, but it is usually
cheaper to concentrate the solution after it has attained a
strength of about 20 per cent., than to overcome too great
a resistance of the solution. The initial cost of the cells is
high and a large amount of mercury is constantly in use.
About 7 tons of mercury are required for each ton of caus-
tic soda produced in a day. The power to move the cell is
small, but must be considered in estimating the cost of
working the plant, and it also adds to the complication
of the plant.

Various estimates have been made of the cost of bleach
and caustic by the electrolytic process, and practically all of
them show that they cost more by this method than by the
older processes. Nevertheless, the electrolytic processes
are able to continue and pay dividends, so that apparently
something is wrong with the calculations. The truth of the
matter is that sodium hydrate can be made more cheaply
by the ammonia-soda process than by any other, but this
process cannot produce chlorine. The electrolytic process
can produce chlorine more cheaply than the Le Blanc proc-
ess, so that we must consider the electrolytic processes as

82 ALKALIES AND HYDROCHLORIC ACID § 30

essentially processes for the production of chlorine and the
caustic soda as a valuable by-product.

108. Electrolytic Bleacb. — One of the main things
■ that we have been struggling against so far has been the
formation of hypochlorite in solution, but nevertheless,
when a bleaching solution is wanted, the hypochlorites are
just what are needed. As early as 1883 Hermite patented
a process and advocated the use of electrolytic bleach. He

proposed to electrolyze solutions of calcium chloride, mag-
nesium chloride, or a mixture of one or both of these with
salt in such a way as to obtain hypochlorites in solution.
This is easily accomplished by placing the cathode over the
anode, so that the chlorine, in rising, must pass through the
caustic formed; and if the electrolyte is kept circulating

§ 30 ALKALIES AND HYDROCHLORIC ACID 83

through the bleach vat, the apparatus lasts well and the
process is satisfactory.

A very satisfactory apparatus for carrying out an electrol-
ysis of this character has been invented by Kellner. Fig. 25
shows an apparatus of this character in vertical section and
ground plan. It consists of the cell c with a cover d. The
side walls, which act as insulators, carry electrode plates ^, e\
etc. and f^ f\ etc. of carbon, or metal with platinum on
one side, which extend alternately into the cell, so that the
electrolyte is forced to zigzag between them in passing from
one end of the cell to the other. The first and last plates
extend through the cover and serve for connecting with the
current. This arrangement makes it possible to electrolyze
the solution in a small space, and also enables the operator
to use a current of high voltage, as is frequently available
from electric-light plants. By regulating the number of
intervening plates, the current can be reduced in voltage for
each section of the cell, in the same manner as would be
the case if a series of the same number of cells were used.
In operation, the electrolyte enters a and flows in the direc-
tion of the current, finally leaving at b. The circulation of
the brine i^ so regulated that about .05 per cent, of active
chlorine is formed at each passage of the brine through the
apparatus. When the brine has 1 per cent, of active chlo-
rine, it is used for bleaching. The composition of the elec-
trolyzed brine depends on the voltage, amperage, tempera-
ture, and the amount of sodium chloride present. The
bleaching solution is clear, has an apple-like, odor, and keeps
better than a solution of bleaching powder having the same
amount of available chlorine.

The question as to whether it will pay to use this method
or not is one that every user of bleach must decide
from the conditions prevailing at his factory. In most
cases it is probably better and cheaper to allow the brine
to be electrolyzed at some central plant, where the sodium
can be saved as hydrate, and there to convert the chlo-
rine into bleach, to be shipped to the place where it
is needed. In some places where large quantities of

84 ALKALIES AND HYDROCHLORIC ACID § 30

bleaching liquors are used, however, it will, without doubt,
pay to make it on the spot, thus saving the carriage of
large amounts of inert material in order to get the neces-
sary chlorine.

POTASSIUM CHIiORATE

103. It has been shown that if the products of electrol-
ysis of an alkaline chloride are allowed to combine, the
result is the formation of the hypochlorite. If now the con-
ditions are suitable, the hypochlorite changes to the chlo-
rate. The total result of the electrolysis of potassium
chloride, when the solution is kept cool and the current
density low, is represented by the equations

If the solution is allowed to heat up, however, the potas-
sium hypochlorite goes over into potassium chlorate and
chloride, according to the reaction

ZKCIO = 2Ka + KC/O,

If we omit the intermediate reaction, we have

6K0H+ 3C7/. = 6Ka+ JCC/O, + ZHfi

as the reaction for the formation of potassium chlorate from
potassium hydrate and chlorine. Finally, neglecting all
intermediate steps, we now write, as representing the final
result of electrolyzing a solution of potassium chloride in
such a manner as to give potassium chlorate, the reaction

KCl-\- ZHfi = KCIO, -f 3//.

A glance at this reaction, recalling Faraday's law, will
show that it takes at least six times as much electricity
to make 1 molecule of potassium chlorate as is required to
decompose a molecule of potassium chloride, or 6 molecules
of potassium chloride are decomposed in order to get 1 mole-
cule of potassium chlorate. It will thus seem that there is

§ 30 ALKALIES AND HYDROCHLORIC ACID 85

a great waste of current in this process; but if it is consid-
ered that by the older chemical methods it required 6 atoms
of chlorine to make 1 molecule of potassium chlorate, it
will be seen that as much loss occurs in the older methods
as in the electrolytic process. The greatest argument in
favor of the electrolytic method is, however, that it has
been running for several years and at a profit, so that appar-
ently it can today make potassium chlorate at least as
cheaply, and probably more cheaply, than it can be made by
the old methods.

That it is possible to make chlorates by electrolysis was
probably first noted by Stadion in 1816; the process was
patented in England by Charles Watt, in 1851, and our
present methods differ only in the details of the process and
in the apparatus.

104. Gall-and-Montlaur Process. — This is the oldest
process by which potassium chlorate has been successfully
manufactured electrolytically. It uses lead-lined, rectan-
gular tanks of about 11,000 gallons capacity and insulated
from the floor by means of oil cups. The same means are
used in insulating the whole building. The anodes are an
alloy of 90 per cent, platinum and 10 per cent, iridium, while
the cathodes consist of a nickel-iron alloy. Large quantities
of hydrogen (about 19,000 cubic feet for each ton of potassium
chlorate) are set free ih the process, and if this comes in contact
with chlorate or hypochlorite, it will reduce it and cause loss.
To avoid the action of the hydrogen, the cathodes are enclosed
in asbestos bags, which aid in carrying off the hydrogen.
About a 25-per-cent. solution of potassium chloride is used
in the electrolysis. This solution must be as pure as possi-
ble, for the presence of metallic oxides causes very rapid
decomposition of the potassium hypochlorite first formed
into potassium chloride and oxygen. An electromotive
force of 5 volts and a current density sufficiently high to
keep the temperature at 50° to 60® C. is employed in the
electrolysis. The relative sizes of the cathode and anode
are so arranged that there is a high current density at the

86 ALKALIES AND HYDROCHLORIC ACID § 30

cathode and a low one at the anode. By this method of
working it is possible to obtain a current efficiency of over
50 per cent. The process is in use at several places.

105. Copbin Ppocess. — This process is quite similar
to the Gall and Montlaur in that it produces the complete
action in the cell. It makes use of secondary electrodes,
however, and causes the electrolyte to circulate between
them. The apparatus consists of cement cells with primary
electrodes at the ends and a large number of platinum
plates set in ebonite frames and placed from 12 to 15 milli-
meters apart, which act as secondary electrodes.

When the current passes, one side of the secondary elec-
trode acts as cathode and the other as anode, and since the
plates are so close together, the reaction between the caus-
tic potash and chlorine takes place readily, and by a high
density current the temperature is kept high enough, so that
the chlorate forms at once. The process is in operation at
Chedde, Savoy, but no details as to its success are available.

106. Blumenbergr Process. — In this process it is
attempted to avoid the secondary decompositions, the high
density, and the reduction by hydrogen by first making caus-
tic potash and chlorine, collecting them separately, and com-
bining them outside of the cell. The potassium chloride is
dissolved, filtered, and run into the storage tanks yl, Fig. 26,
from which it can run directly into the electrolysis cell.
This consists of a simple cell B^ divided into anode c and
cathode d compartments by a simple diaphragm e. During
electrolysis the potassium hydrate collects in the cathode
compartment and the chlorine is saved in the gas holder /^
When the electrolysis has continued long enough to give
considerable caustic, the contents of both compartments
c and d are allowed to mix in the pan (7, and the chlorine is
run in from the gas holder F to form the chlorate. Both the
electrolysis cell and G are arranged so that they can be
heated by means of steam pipes when necessary. From G
the chlorate is run down into concentration and crystalliza-
tion tanks. High efficiency is claimed for this process.

§ 30 ALKALIES AND HYDROCHLORIC ACID 87

107. Glbbe Process.— The fact that during the for-
mation of potassium chlorate by electrolysis large quantities
of hydrogen are formed has already been mentioned. In
order to avoid the reducing action of this gas, Gibbs makes
use of the cathodes of copper oxide so that the hydrogen is
oxidized as rapidly as it is set free. This also reduces the
polarization at the electrodes. In other respects the Gibbs
method has little that is peculiar to it, and this feature

seems to be adapted from the Richardson-and- Holland proc-
ess for caustic and chlorine. In the cell the cathode is
placed above, the anode below, and the temperature is kept
at 80° or 90° C. When one-half of the potassium chloride
in the cell has been converted into chlorate, the liquor is
drawn off and the chlorate crystallizes out. The cathodes
are then renewed, and those just used are reoxidized by
heating in the air. This process is at present in successful
operation at Niagara Falls.

88 ALKALIES AND HYDROCHLORIC ACID § 30

108. It has recently been shown that the presence of
alkaline carbonates or the alkaline-earth hydrates in the
cell greatly increases the yield of the chlorate, and proba-
bly all of the factories use one of these substances or calcium
chloride as a constituent of the cell solution. In just what
way these materials act is at present unknown, although
several theories have been advanced.

ALKALIES AND
HYDROCHLORIC ACID

(PART 3)

ANALYTICAL METHODS

AMMONIA SODA

CRUBE MATERIAI^

1. Brine. — 1. The specific gravity is determined by
means of a hydrometer or specific-gravity spindle; the
amount of salt in the brine can then be stated with a
fair degree of accuracy by reference to a table. This test
is so rapidly made that it is used frequently for checking
the brine as to its salt content when it comes to the works.
Table I gives the percentage of sodium chloride correspond-
ing to each specific gravity from 1 per cent, of salt to a
saturated solution.

For convenience, special hydrometers are frequently used
which are so graduated that percentage of salt is read direct,
or the point where it stands in a saturated salt solution is
marked 100 and the stem between this point and that which
pure water gives is divided into 100 parts, so that the
observer reads the percentage of the saturation of the
brine.

§31

For notice of copyright, see page immediately following the title page.

2 ALKALIES AND HYDROCHLORIC ACID § 31

To obtain the number of grams of salt in a liter of its
brine, we move the decimal point in the specific-gravity
value one point to the right and multiply by the percentage

TABIiE I

Specific

Per cent.

Specific

Per cent.

Specific

Per cent.

Gravity

NaCl

Gravity

NaCl

Gravity

NaCl

1.00725

1.0733s

1. 14315

19.000

1. 01450

1.08097

1. 15107

20.000

1. 02174

1.08859

1.15931

21.000

1.02899

1.09622

1. 16755

22.000

1.03624

1. 10384

1. 17580

23.000

1.04366

1:11146

1. 18404

24. 000

1. 05108

1.11938

1. 19228

25.000

1. 05851

1. 12730

1.20098

26.000

1.06593

1.13523

1.20433

26.39s

of salt at the specific gravity observed. For example, if the
specific gravity is 1.204, then 12.04 X 26.39 = 317.74 grams
per liter.

2. Inorganic sediment is determined by filtering 500 cubic
centimeters of the brine through a filter of known ash,
igniting, and weighing. After subtracting the ash and
multiplying the remainder by 2, the result is grams of
inorganic sediment per liter of brine.

3. Ferric Oxide and Alumina, — 200 cubic centimeters
of filtered brine is acidified with a few cubic centimeters of
nitric acid and heated for 10 minutes in order to oxidize any
possible ferrous compounds, made slightly alkaline with
ammonium hydrate, warmed 10 minutes, and filtered.
The precipitate is redissolved in hydrochloric acid and the
ferric oxide and alumina determined as usual. The number
of grams of Fefi^ and Alfi^ X 5 = grams of Fefi^ and
Alfi^ per liter of brine.

§ 31 ALKALIES AND HYDROCHLORIC ACID 3

4. Calcium oxide is determined in the filtrate from the
iron and alumina. One-half gram of ammonium chloride is
added, and to the hot ammoniacal solution sufficient ammo-
nium oxalate added to precipitate all the calcium. The
calcium oxalate is filtered off, strongly ignited over a blast,
and weighed as CaO, This weight, after subtracting the
filter ash and multiplying by 5, gives the grams of CaO per
liter of brine.

5. Magnesia is determined in the filtrate from the cal-
cium precipitate by adding ammonium phosphate and strong
ammonia solution, equal to one-third of the total volume of
the solution, allowing to stand 24 hours, filtering, washing
with dilute ammonia water, igniting at red heat, and weigh-
ing. The precipitate, after subtracting the filter ash, is
magnesium pyrophosphate Mg^Pfi^, Wt. Mg^Pfi^ X .36036
X 5 = grams MgO per liter of brine.

6. Sulphur Trioxide. — 60 cubic centimeters of the
filtered brine is acidified with a few drops of hydrochloric
acid, diluted with an equal volume of distilled water, and
heated to boiling. Boiling hot barium chloride is slowly
added in slight excess and the whole allowed to stand until
the precipitate completely settles. The barium sulphate is
then filtered off and washed with hot water, first by decan-
tation and then on the filter until free from chlorides.
The precipitate is then ignited and weighed as usual. The
weight of barium sulphate BaSO^ X .34335 X 20 = grams
SO^ per liter of brine.

7. Sodium Chloride, — The amount of salt in the brine is
usually determined with sufficient accuracy by means of the
hydrometer. If a more accurate determination is wanted,
10 cubic centimeters of the clear brine is diluted to
1,000 cubic centimeters and 10 cubic centimeters of this
dilute solution is titrated with -^ normal solution of silver
nitrate, using potassium chromate as indicator. The num-
ber of cubic centimeters of -j^ normal solution of silver
nitrate X .00355 X 10,000 = grams of chlorine per liter.

4 ALKALIES AND HYDROCHLORIC ACID § 31

Or, without a very great error, we may state, number cubic
centimeters -^ normal AgNO^ solution X .00585 X 10,000
= grams salt per liter.

2. Groi^pingr of Substances Determined. — The most
rational method of procedure is to report each substance as
found, but it is a very common requirement that the results
shall be reported grouped together so as to form salts. In
this case, this result is obtained by combining the SO^ and
CaO to form CaSO^y any excess of calcium and the mag-
nesium are then combined with chlorine and the excess of
chlorine is then calculated as salt. Ferric oxide and
alumina are usually reported as such.

3. Limestone. — For the analysis of an average sample
of the limestone used through the month, and, in general,
for a careful control of the materials used, the method of
analysis given for limestone in Quantitative Analysis
should be used. It frequently happens, however, that it is
necessary to analyze one or more samples of the rock each
day as it comes from the quarries. In that case the following
more brief method is preferable.

1. Insoluble, — 1 gram of the limestone is treated with
an excess of dilute hydrochloric acid, warmed, filtered,
washed, ignited, and weighed. In case the limestone con-
tains a large amount of organic matter, this may be deter-
mined by filtering through a filter paper that has been
pleated to 100° C, cooled in a desiccator, and weighed. In
this case the insoluble matter is dried at 100° C, cooled,
and weighed before ignition. The difference between the
weight of the insoluble matter before and after ignition
gives the amount of organic insoluble matter.

2. Lime, — Dissolve 1 gram of the sample in 25 cubic
centimeters of normal hydrochloric acid and titrate back to
the neutral point with normal soda solution, using methyl
orange as indicator. The difference between the number of
cubic centimeters of acid and alkali used gives the number

g 31 ALKALIES AND HYDROCHLORIC ACID 5

of cubic centimeters of acid neutralized by the limestone.
The number of cubic centimeters of acid used x 2.8 = per-
centage of CaO; or number of cubic centimeters of acid
used X 5 = percentage of CaCO^, By this method the
magnesium carbonate in the limestone is reported as a cal-
cium compound, but for most limestone used in the ammonia-
soda industry this can be overlooked.

3. Magnesia, — In case the amount of magnesia is
required, dissolve 2 grams of limestone in hydrochloric acid
and precipitate the calcium directly by the addition of ammo-
nium hydrate and ammonium oxalate to the hot solution.
Allow to stand 10 minutes at a gentle heat, then filter and
wash. Determine the magnesium in the filtrate by pre-
cipitating with ammonium phosphate. Make the solution
strongly ammoniacal and let stand 2 hours with frequent,
thorough stirring. The precipitate may then be filtered
off, ignited, and weighed as Mg^Pfi^. ^Jg^Pfi^ X 18.018
= percentage of MgO in the limestone; Mg^P^O^ X 37.808
= percentage of MgCO^ in the limestone.

We can now correct the value obtained for lime. The
molecular weight of MgO = ^^ and of CVi(7=56, there-
fore CaO is \%, or 1.4 times heavier than MgO\ therefore
the percentage of MgO X 1.4 = percentage of CaO that this
percentage of magnesia would give as lime. In the same
way we find that the percentage of MgCO^ X 1.19 = percent-
age of CaCO^. It is now possible to report the percentage of
lime or calcium carbonate in the limestone with a fair degree
of accuracy. For example, if it is found that the limestone
apparently contains 95 per cent, of CaCO^ and then find
2 per cent, of MgCO^^ the true percentage of CaCO^ is 95
— (2 X 1.19) = 92.62 per cent.

4, Quicklime. — The analysis of the monthly average
and the careful check determinations should be carried out
in the same manner as is described for limestone in Quanti-
tative Analysis, In reporting the result of the analysis, the
carbon dioxide and sulphur trioxide are combined with the
lime, and the remainder of the calcium and the magnesium

6 ALKALIES AND HYDROCHLORIC ACID § 31

reported as oxides. Where less accurate results will answer,
the following method is preferred.

1. Insoluble, — The amount of insoluble matter is deter-
mined as in Art. 3,

2. Free Calcium Oxide. — Weigh out 50 grams of an aver-
age sample of the lime, and after carefully slaking it, bring
the mass into a 1,000-cubic-centimeter measuring flask, fill
to the mark, and thoroughly mix. Pipette out, without
allowing the suspended matter to settle, 100 cubic centi-
meters and dilute to 500 cubic centimeters, shake thoroughly
and pipette out 100 cubic centimeters for titration with
normal hydrochloric acid, using phenol phthalein as indi-
cator. The number of cubic centimeters of acid required
to just discharge the pink color multiplied by 2.8, gives
the percentage of CaO,

3. Calcium Carbonate, — Titrate 1 gram of the sample,
using methyl orange as indicator, as under Art. 3, By sub-
tracting the number of cubic centimeters of normal acid
required above from the number of cubic centimeters
required here, the number of cubic centimeters of normal
acid required for the calcium carbonate is obtained. This
value multiplied by 5 gives the percentage of calcium car-
bonate.

4. Magnesia. — Determine as under Art. 3.

5. Ammonia lilquor. — The crude ammonia liquor as it
comes to the works from the gas manufacturer frequently,
especially in cold weather, contains crystals. The liquor is
measured and the crystals weighed before sending them to
the storage tanks, and a sample of each is sent to the labo-
ratory for analysis.

1. Specific Gravity. — The specific gravity of the gas
liquor is taken with a hydrometer. This determination is,
however, of very secondary importance to the direct deter-
mination of the ammonia.

§ 31 ALKALIES AND HYDROCHLORIC ACID 7

2. Ammonia is determined, both in the crystals and in
the gas liquor, according to the volumetric method described
in Quantitative Analysis,

6. Coal and Ck>ke. — These are analyzed by the method
described in Quantitative Analysis.

INTEBMBDIATE PRODUCTS

7. Ammoniacal Brine* — The determinations ordinarily
made are free and combined ammonia and salt. i

1. Free and Combined Ammonia, — Dilute 10 cubic centi-
meters of the ammoniacal brine with distilled water to about
100 cubic centimeters, introduce it into a distilling flask (see
Quantitative Analysis) , and boil until all the ammonia and
ammonium carbonate are driven off. The ammonia and
ammonium carbonate are collected in normal sulphuric acid
and determined as usual. The result is free ammonia.

A new receiver containing normal sulphuric acid is then
attached and ammonia-free sodium-hydrate solution is intro-
duced into the distilling flask. The combined ammonia
is then driven over by the boiling and is determined by
titrating the acid in the receiver.

2. Salt. — On account of the free alkali present in this
brine the common method of titrating with silver nitrate
cannot be used, unless the ammonia is exactly neutralized
with nitric acid ; even then the results lack exactness. The
so-called Volhard method, which possesses the advantage
that it can be used in a nitric-acid solution, is therefore
used. This method is described in Quantitative Analysis,

8, lilme-Klln Gases. — Carbon dioxide, carbon mon-
oxide, and oxygen must be determined in the gases coming
from the lime kiln. These determinations may be made with
the Orsat-Muenke apparatus, described in Quantitative
Analysis^ under ** Gas Analysis," or, on account of its cheap-
ness, by means of the Bunte burette.

8 ALKALIES AND HYDROCHLORIC ACID g 31

0. Bunte Burette. — The apparatus shown at e, Fig. 1,
consists of a simple glass tube a little over 100 cubic
centimeters capacity and closed at each end by well-
fitting stop-cocks ^ and /,
The stop-cock at g- is the
ordinary two-way style; the
one at _/ is a three-way stop-
cock, so that the tube can be
put in connection with the
source of gas through the end
of the cock and the rubber
tube t/i, or it can be connected
with the cup-shaped recepta-
cle /, which is made above /".
The tube e is graduated in
-jV cubic centimeters for
100 cubic centimeters down
from /. Frequently the bu-
rette is surrounded by a
water-jacket to prevent vari-
ations of temperature. This
is, however, an unnecessary
accessory and seriously inter-
feres with the manipulation
of the burette.

10. Manipulation of the
■'"'■' Bunte Burette. — The

burette is first filled with gas to be analyzed.

{a) If only a limited amount of gas is available for
analysis, the burette is first filled with water from the reser-
voir ^, Fig. 1. For this purpose the rubber tuber is attached
to the tip of the burette ;/, the stop-cock^ is opened, and /is
turned so as to connect the burette with ;«. Then by releas-
ing the pinch cock rf water flows from d to fill the whole
apparatus. The stop-cocks are then closed, c is removed
from «, and m is attached to the gas supply. By then again
opening the stop-cocks /"and ^ the water flows out at « and

§ 31 ALKALIES AND HYDROCHLORIC ACID 9

the burette fills with gas, which is then secured by closing
the stop-cocks.

{d) When gas is abundant and under pressure, the tube ;//
is attached to the source, / and ^ are opened, and 2 or 3 liters
of gas allowed to flow through the burette, thus sweeping
out the air and leaving a good sample of gas. By closing /
and ^ the gas is enclosed.

(c) When the gas is abundant, but not under pressure,
as happens in taking samples between the lime kilns and the
pumps, it is necessary to attach an aspirator at ;/ to draw
the gas through the burette. A suitable arrangement for
aspirating in this case consists of a large bottle o. This
bottle is filled with water, the rubber tube p attached at ;/,
the pinch cock r opened, and then the stop-cocks g and /
opened. After 2 or 3 liters of gas have been drawn through
the burette, first g" and then / is closed, p is disconnected
from «, and the sample is ready for analysis.

Having the burette filled with the sample of gas, the cup /
is filled with water to a mark that is 1 centimeter above the
stop-cock y, c is then attached to ;/, and d and g" opened.
Water thus flows into the burette and compresses the gas.
When the water reaches the 100-cubic-centimeter mark, ^is
closed and /"is turned to connect the burette with /. Gas
will escape until the gas in the burette is under the atmos-
pheric pressure, plus the pressure of 1 centimeter of water.
f is then closed and the volume of gas read (it should be
exactly 100 cubic centimeters). The rubber tube i of the
suction flask A is then attached at //, g" is opened, and, by
sucking on j\ the water is almost completely removed from
the burette, leaving a partial vacuum; ^ is then closed, and /
removed from n.

For the determination of carbon dioxide, a small beaker
containing a suitable solution of caustic potash is brought
under n, and ^ turned so that the alkali solution rises in the
burette; g is then closed. The burette is then grasped at /,
loosened from the clamp k (the ends of the burette are
grasped between the first and second fingers of each hand

10 ALKALIES AND HYDROCHLORIC ACID § 31

beyond the stop-cocks to avoid heating the gas by the
hands), and after the water is emptied from /, the burette is
thoroughly shaken, so that the gas is well mixed with the
caustic potash. The burette is then replaced in the clamp >fe,
n is brought under caustic-potash solution, and g again
opened. The alkali will rise in the tube, and when it has
filled as much as it will, g is once more closed and the burette
shaken as before. This is repeated as long as the alkali
solution continues to rise in the burette. Water is then
filled to the 1-centimeter mark in /, / is opened to insure
equal pressure, then closed, and the volume of gas read.
The difference between this reading and 100 gives the volume
percentage of the carbon dioxide in the gas mixture.

For the determination of oxygen, the caustic potash is
removed as far as possible by means of the suction flask ^,
and alkaline pyrogallol allowed to rise in the burette in its
place. The same operations as for carbon dioxide are per-
formed until all the oxygen is absorbed. The volume of
gas is then read. The difference between this volume and
100 gives the volume percentage of carbon dioxide and oxy-
gen, and deducting the volume percentage of carbon diox-
ide leaves the volume percentage of oxygen in the gas.

In each of the above cases the gas is read over strongly
alkaline liquids that tend to adhere to the burette and ren-
der the results inaccurate. This can be avoided by sucking
out the alkaline liquid, allowing water to enter, rinsing the
burette two or three times, each time sucking out the
water, and then measuring the gas over nearly pure water.

The carbon monoxide is determined by sucking out the
alkaline pyrogallol or water after measuring the oxygen,
replacing it with a hydrochloric-acid solution of cuprous
chloride, and proceeding as in the preceding cases. After
the carbon monoxide has been completely absorbed, as shown
by the absorbing liquid no longer rising in the burette, the
absorbing liquid is sucked out as completely as possible and
the gas washed two or three times with water to completely
remove the hydrochloric acid. This diminution in volume
of the gas gives the volume percentage pf carbon monoxide

§ 31 ALKALIES AND HYDROCHLORIC ACID 11

in the gas; the remainder of the gas is the volume percent-
age of nitrogen in the gas.

11, Reagents for the Bonte Burette* — The caustic
potash is made by dissolving 100 grams solid potassium
hydrate in 200 cubic centimeters of water.

The alkaline pyrogallol is made by dissolving 32 grams
potassium hydrate in 200 cubic centimeters of water and
40 grams of pyrogallic acid in 200 cubic centimeters of water.
The two solutions are thoroughly mixed and kept carefully
guarded from the air in a rubber-stoppered bottle. It is
even better to keep the two solutions separate and only mix
them when needed for use.

The cuprous-chloride solution is made by dissolving
200 grams of cupric chloride in 500 cubic centimeters of water
and 500 cubic centimeters of concentrated hydrochloric acid,
and allowing the solution to stand tightly stoppered in a
bottle containing copper turnings or strips of sheet copper
until it becomes clear and colorless.

•

13, lilquor From Carbonators. — The free and com-
bined ammonia are determined as described in Art. 7. These
are the only determinations usually made.

13. Bicarbonate From the Filters. — 1. Total alkali is
determined by titrating 4.2 grams of the sample with normal
sulphuric acid, using methyl orange as indicator. Each
cubic centimeter of normal acid used corresponds to. 738 per
cent. oiNafi in the sample.

2. Sodium Bicarbonate. — The determination of sodium
bicarbonate in the presence of sodium carbonate depends
on the reaction

NaHCO, -t- NaOH= Na.CO, + H^O

Silver nitrate is used as indicator, for it gives a white pre-
cipitate with sodium carbonate, but as soon as a single drop
of caustic-soda solution is present in excess the silver car-
bonate precipitate turns brown, owing to the formation of
silver oxide.

la ALKALIES AND HYDROCHLORIC ACID § 31

Normal sodium-hydrate solution is prepared by dissolv-
ing 50 grams of pure sodium hydrate in 1 liter ol water and
adding sufficient barium hydrate to more than precipitate
all the carbon dioxide. The solu-
tion is then standardized as usual
by titrating with normal sul-
phuric acid, using phenol phthal-
ein as indicator, and then cor-
rected to exactly normal strength.
This solution must after stand-
ardization be carefully guarded
from the carbon dioxide of the
air.

A convenient arrangement for

the solution and burette is shown

in Fig. a. The burette a is

closed at the top with a stopper,

through which passes a glass tube

connecting with a sugar funnel ^,

which is filled with pieces of soda

lime and so removes the carbon

dioxide from the air that enters

the burette. At the lower end of

the burette a tube is blown on

'"■ which connects, by means of the

glass tube f/and two short pieces of rubber tube, with the

bottle e containing the standard solution. The bottle f is

closed with a two-holed rubber stopper, through one hole of

which leads the tube </to the burette, and through the other

a glass tube to the sugar funnel / that contains the soda

lime. The liquid can be started first by blowing on the end

of / after the stop-cock c has been opened. After the

apparatus is once in operation the burette can be repeatedly

filled, by merely opening the stop-cock c, without exposing

the solution to the air at any point.

The determination iS made by weighing out in a beaker
4.3 grams of the sample, adding lOO cubic centimeters of water
(not warmer than 20° C), and running in the caustic-soda

§ 31 ALKALIES AND HYDROCHLORIC ACID 13

solution until within about 1 cubic centimeter of the end
reaction. The solution is then thoroughly stirred and the
standard solution run in, at first .2, and then .1 cubic centi-
meter at a time, until a drop taken out and brought in con-
tact with a 25-per-cent. silver-nitrate solution on a white
plate shows a brown color at once. Even before the end
point, the drops turn brown on standing. If the compo-
sition of the sample is not approximately known at first, it
must be approximately determined by weighing out a por-
tion of the sample and running in the standard caustic 2 or
3 cubic centimeters at a time and testing until the end
point is passed. Then, for the final determination, some-
what less than this amount of the standard solution is taken
as above. The number of cubic centimeters of the normal
alkali used multiplied by 2 gives the percentage of sodium
bicarbonate in the sample.

3. The percentage of sodium carbonate in the sample is
given by multiplying the difference between the number of
cubic centimeters of normal acid required for the total alkali
and the number of cubic centimeters of normal caustic alkali
required for sodium bicarbonate by |f . For example, if it
takes 39 cubic centimeters of normal acid to neutralize a
sample and 35 cubic centimeters of normal alkali to convert
the bicarbonate into the carbonate, then 39 X .738 = 28.79 per
cent, of Nafi\ 35 X 2 = 70 per cent, of sodium bicarbonate;
and (39 — 35) x f J = 5.05 per cent, of sodium carbonate.

4. Ammonia is determined according to the volumetric
method given in Quantitative Analysis,

5. Moisture is determined by weighing out 10 grams of
the sample in a small platinum or porcelain evaporating
dish and heating, at first carefully on a sheet of asbestos,
and finally to from 300° to 400° C. The loss in weight, -
after deducting the carbon dioxide corresponding to the
sodium bicarbonate, gives the moisture.

14, Mother lilquor. — The mother liquor from the filtra-
tion of the liquors from the carbonators is tested for free
and combined ammonia and salt.

14 ALKALIES AND HYDROCHLORIC ACID § 31

1. Free and combined ammonia are determined as under
Art. 7.

2. Salt is determined by evaporating 10 cubic centi-
meters of the liquor to dryness in a platinum dish, heating
the residue until the ammonium chloride is volatilized, then
cooling and weighing.

15. . Milk of liime. — 1. The determination of the specific
gravity usually is sufficient for controlling the milk of lime.
If the milk of lime is thin, it is thoroughly mixed and the
reading on the hydrometer is quickly taken. If the milk of
lime is thick, a rather broad cylinder is selected, the milk
of lime thoroughly mixed, the hydrometer inserted, and
the cylinder jarred on the table until the hydrometer will
sink no lower, when it is read. A hydrometer called the
Baumi hydrometer^ with the spindle arbitrarily divided
•into so-called degrees, is frequently used for this purpose.
Table II shows the degrees Baum6 and grams per liter of
calcium oxide corresponding to a considerable range of
specific gravities.

2. Complete Analysis, — At intervals a complete analysis
of the milk of lime is required. For this purpose the sample
is thoroughly mixed, and 250 cubic centimeters measured
out and filtered. The residue on the filter is taken without
washing, dried at 100° C, and weighed. This weight mul-
tiplied by 4 gives the undissolved portion per liter.

The undissolved portion and the filtrate are then sepa-
rately analyzed, exactly as under ** Quicklime."

16, Waste From Ammonia Stills. — 1. Excess of lime
is the constituent of this waste, concerning which it is most
important for us to have information — that is, the lime
that is still available for liberating ammonia from its salts.
For its determination, boil 100 cubic centimeters of the
waste until no more ammonia is given off, then add
ammonium sulphate in excess, boil again, and collect the
ammonia evolved this time in normal acid (see the volumetric
determination of ammonia. Quantitative Analysis). By
titrating, the necessary ihformation for finding the amount

§ 31 ALKALIES AND HYDROCHLORIC ACID 15

of ammonia evolved rs obtained, and from this it is a simple
matter to calculate the amount of free lime in the waste.

{N/fX^O, +

Ca{OH\

%NH.

' = CaSO, + -^^ + Hfi
74 34

34 : 74 = wt. NH^ found : x
jr X 10 = the amount of available lime per liter of the waste.

TABIiE n

Specific

Degrees

Grams CaO

Specific

Degrees

Grams CaO

Gravity

Baum6

in Liters

Gravity

Baum6

in Liters

1.007

7.5

1. 125

159

1. 014

16.5

1. 134

170

1.022

26.0

1. 142

181

1.029

36.0

1. 152

193

1.037

46.0

1. 162

206

1.045

56.0

1. 171

218

1.052

65.0

1. 180

229

1.060

75.0

1. 190

242

1.067

84.0

1.200

255

1.075

94.0

1. 210

268

1.083

104.0

1.220

281

1. 091

115. 0

1.231

295

1. 100

126.0

1. 241

309

1. 108

137.0

1.252

324

1. 116

148.0

1.263

339

2. Complete Analysis, — Determine the specific gravity,
the amount of undissolved material, and analyze the insoK
uble portion as in Art. 15, In the soluble portion:

(a) Titrate 50 cubic centimeters with normal sulphuric
acid, using phenol phthalein as indicator, and calculate the
result as Ca(OH)^.

{b) Determine the calcium in 25 cubic centimeters, as
usual, by precipitating with ammonia and ammonium oxa-
late, filtering, and titrating the precipitate with potassium

16 ALKALIES AND HYDROCHLORIC ACID § 31

permanganate. Deduct the calcium corresponding to the
amount of calcium hydrate found under {a) and calculate
the remainder as calcium chloride in grams per liter.

{c) Determine the sulphur trioxide in 50 cubic centi-
meters by precipitating with barium chloride, and calculate
the result as sodium sulphate in grams per liter.

(d) Determine the chlorine in 5 cubic centimeters by
Volhard*s method. Deduct the chlorine corresponding to
the calcium chloride found under {d) and calculate the
remainder as sodium chloride in grams per liter.

THE FINISHED PKODUCT

17, Soda Ash. — For the complete analysis of soda ash,
the following determinations are usually made :

1. Sodium Carbonate, — Weigh out 2.65 grams of the dry
substance, dissolve in about 150 cubic centimeters of water,
and titrate with normal sulphuric acid, using methyl orange
as indicator. The number of cubic centimeters of acid used
multiplied by 2 gives the percentage of sodium carbonate.

2. Sodium Bicarbonate. — This substance rarely occurs in
large amounts in soda ash, and its determination may
usually be omitted. If there is a reason for determining it,
use the method given under Art. 13,

3. Sodinm Chloride, — Dissolve 5 grams of the sample in
water and titrate by Volhard's method.

4. Silica, — Dissolve 50 grams of the sample in about
150 cubic centimeters of water and acidify with concen-
trated hydrochloric acid, evaporate to dryness on the water
bath, take up with water and a little hydrochloric acid,
filter, ignite, and weigh. Calculate as silica; of course it
consists of everything insoluble in hydrochloric acid.

5. Ferric Oxide and Alumina. — Determine the ferric
oxide and alumina in the filtrate from the silica by precipi-
tating with ammonia as usual.

6. Calcium Carbonate, — Divide the filtrate from the
above determination into two equal parts, and in one half

§ 31 ALKALIES AND HYDROCHLORIC ACID 17

determine the calcium, as usual, with ammonia and ammo-
nium oxalate, and calculate as calcium carbonate.

7. Magnesium Carbonate. — Determine the magnesium in
the filtrate from the calcium determination, as usual, with
ammonium phosphate, and calculate as magnesium car-
bonate.

8. Sodium Sulphate, — Determine the sulphur trioxide in
the other half of the filtrate from the ferric oxide and
alumina determination by means of barium chloride, as
usual, and calculate as sodium sulphate.

A complete analysis of this character is necessary from
time to time, usually each month, of an average of the soda
ash made. For the daily control of the output, however,
a determination of the sodium carbonate and the sodium
chloride is generally sufficient.

SAIiT-CAKK PROCESS

CRUDE MATBRIAIiS

18. Salt. — The usual determinations are as follows:
1. Sodium Chloride, — Weigh out 4 grams of the sample, dis-
solve in water, and dilute to 1,000 cubic centimeters. Take
50 cubic centimeters of this solution and titrate with ^ nor-
mal silver nitrate, using about \ cubic centimeter of potas-
sium chromate as indicator. This gives the total chlorine,
and when no other substances are determined, this is all
calculated as sodium chloride. When magnesium and other
substances present as chlorides are determined, the chlorine
of these is first subtracted from the total before calculating
it as sodium chloride.

2. Water, — The determination of water in salt offers
some difficulties on account of its tendency to decrepitate
and so fly out of the dish in which one is heating it. The
most satisfactory method of making the determination is to
select a tall Erlenmeyer flask of Jena glass, of about
250 cubic centimeters capacity, and weigh it with a small

18 ALKALIES AND HYDROCHLORIC ACID § 31

•

funnel in its mouth. About 5 grams of salt are then intro-
duced and its weight exactly established by weighing flask,
funnel, and salt. The funnel is then removed and the flask
is heated for 3 or 4 hours on a suitable sand bath, which has
a temperature of about 150° C. The funnel is then replaced
in the mouth of the flask and the whole allowed to cool and
then weighed. The funnel serves the purpose of preventing
the air from circulating in the flask, so it can be cooled out
of a desiccator. This determination gives all the water in
the salt except part of that which is chemically combined
with impurities. For most purposes the combined water
can be neglected, but when it is necessary to determine it,
this can be done by heating the flask to 300° or 400° C. with
the funnel in its mouth, cooling and weighing.

3. Sulphur Trioxide, — Dissolve 10 grams of salt in
about 300 cubic centimeters of water, acidify with hydro-
chloric acid, and digest at 70° or 80° C. for an hour to
dissolve all the calcium sulphate present. Make this to
600 cubic centimeters, filter through a dry filter, and take
250 cubic centimeters for analysis. Determine the sulphur
trioxide by precipitating,- as usual, with barium chloride in a
hot solution. Unless there are reasons for doing otherwise,
the sulphur trioxide is calculated as calcium sulphate.

4. Other Determinations, — These determinations are
sufficient for the daily work, unless salt happens to come in
from a new source, when it must be analyzed like the aver-
age sample below. The daily samples are saved, however,
and at the end of each month an average sample is prepared
and, in addition to the above determinations, insoluble in
acids, ferric oxide and alumina, calcium, and magnesium are
determined. For this purpose 50 grams of the sample are
dissolved in water and hydrochloric acid and the determi-
nations are carried out as under Art. !•

The magnesium is calculated as chloride, and the calcium
in excess of the sulphur trioxide is calculated as calcium
chloride. Conversely, any sulphur trioxide in excess of the
calcium is calculated as sodium sulphate.

§ 31 ALKALIES AND HYDROCHLORIC ACID 19

FTNISHED PRODUCT

19, Salt Cake. — The determinations usually made are
as follows:

1. Free Acid, — Dissolve 20 grams of the salt cake in
water and dilute to 250 cubic centimeters. Take 50 cubic
centimeters and titrate with normal sodium-hydrate solu-
tion, using methyl orange as indicator. The acidity is
calculated as sulphur trioxide, although it may be due
to hydrochloric acid and salts of the heavy metals, as
well as acid sodium sulphate. If the salt cake contains
large amounts of iron and aluminum salts, and it is desired to
exclude the acidity due to these salts, the titration may be
carried on without an indicator and the end point taken
when flakes of the precipitate of the hydrates begin to
appear. Each cubic centimeter of sodium-hydrate solution
used corresponds to 1 per cent, of sulphur trioxide.

2. Salt, — Take 50 cubic centimeters of the solution pre-
pared as above and determine the chlorine according to
Volhard's method, using -^ normal silver nitrate. Calcu-
late all the chlorine to sodium chloride. Each cubic centi-
meter of silver-nitrate solution used corresponds to .0731 per
cent, of salt.

For the daily determinations, these two substances are all
that are necessary, except when the salt cake is being made
especially free from iron for use in glass manufacture, when
this must also be determined in each batch. For the monthly
average sample and for certain cases for shipment, it is also
necessary to make the following determinations :

3. Insoluble in Acids, — Determine in 50 grams of sample,
as under ** Silica," Art. 17.

4. Ferric Oxide, — Weigh out 20 grams of the sample,
reduce with zinc and sulphuric acid, and titrate with per-
manganate, as directed in Quantitative Analysis,

5. Alumina, — Dissolve 20 grams of the sample in about
150 cubic centimeters of water, add hydrochloric acid, and

20 ALKALIES AND HYDROCHLORIC ACID § 31

precipitate with ammonia as usual. After weighing the
combined oxides, deduct the ferric oxide found above and
calculate the remainder to the percentage of alumina.

6. Lime, — Determine, as usual, in the filtrate from the
alumina determination.

7. Magnesia. — Determine, as usual, in the filtrate from
the lime determination.

8. Sodium Sulphate. — The determination of the sodium
sulphate in this case is a rather difficult matter and it is fre-
quently taken as the difference between the total percentage
of the other substances found and 100. Perhaps the most
satisfactory method of procedure is to dissolve 2 grams of
the sample in as little hot water as possible, make alkaline
with ammonia, and precipitate so far as possible with
ammonium carbonate. Filter and redissolve the precipitate
in as little hydrochloric acid as possible and reprecipitate
with ammonia and ammonium carbonate. Filter and unite
the two filtrates in a platinum dish and evaporate to dryness,
moisten the residue with sulphuric acid to be certain that
the salt present is all converted into sulphate, heat to drive
off the excess of acid, and weigh. Calculate the salt found
by Volhard's method to sulphate, deduct this weight from
that found above, and the remainder is sodium sulphate.

liB BliANC PROCESS

CRUDE MATERIALS

30. Salt cake is analyzed according to Art. 19.

21. lilmestone is analyzed according to Art. 3.

33. Coal is analyzed according to the method given in
Quantitative Analysis. In addition, determine the nitrogen
by Kjeldahl's method, which also is described in Quantita-
tive Analysis,

§ 31 ALKALIES AND HYDROCHLORIC ACID 21

INTERMEDIATE PBODUCrrS

33« Black Asli. — The obtaining of a representative
sample presents perhaps more difficulties than are usually
the case, for the charges as drawn from the furnace are hard
and very non-homogeneous, so that great care must be
exercised in selecting the sample to get it as representative
as possible, for even at best it is imperfect. After the sam-
ple has been carefully selected, it is rapidly crushed and
mixed so that 50 grams of an average of the sample can be
weighed out. These 50 grams are rapidly but thoroughly
ground in a mortar and then brought into a 500-cubic-centi-
meter flask, the mortar rinsed down with water, which has
been boiled to expel carbon dioxide, and then cooled to about
35° C. The rinsings of the mortar are poured into the flask
and the flask filled nearly to the 500-cubic-centimeter mark
with the same warm water. During the pouring of the
rinsings and water on the black ash, it must be thoroughly
shaken to prevent its caking together on the bottom of the
flask. The flask is then allowed to stand about 2 hours with
frequent shaking. A preferable arrangement, and one that
saves much work, is to use one of the many stirrers that run
by a turbine or an electric motor. They may be obtained
from any dealer in chemical apparatus. After standing
2 hours the flask is filled to the mark and the solution is
ready for use.

1. Free Lime, — Thoroughly mix the contents of the flask
and pipette out 25 cubic centinieters of its contents into a
porcelain dish. The outside of the pipette should be rinsed
off before running out its contents and then the inside
should be rinsed into the porcelain dish. Add an excess of
a 10-per-cent. barium-chloride solution and titrate with
normal hydrochloric acid, using phenol phthalein as indica-
tor. Each cubic centimeter of acid solution equals 1.12 per
cent, of CaO,

2. Total Lime, — Pipette out, as above, 25 cubic centi-
meters from the supply flask into a small flask, make acid
with concentrated hydrochloric acid, and boil to expel all

22 ALKALIES AND HYDROCHLORIC ACID § 31

the carbon dioxide. Add a few drops of methyl orange and
then sodium carbonate to exactly neutralize. Add 40 cubic
centimeters of a normal sodium-carbonate solution and boil
to precipitate all the calcium (together with magnesium, etc.,
which, however, can be neglected) as the granular carbon-
ate. Make up to 260 cubic centimeters and filter through
a dry filter. Take 126 cubic centimeters and titrate back
to neutral with normal hydrochloric acid, using methyl
orange as indicator. Each cubic centimeter of the sodium
carbonate used in excess of the acid required to titrate back
is equal to 2.24 per cent, of CaO. Neither of the above
methods is very exact, but they answer for factory con-
trol. The supply flask is now tightly stoppered and allowed
to stand until the liquor has become completely clear.

3. Total alkali comprises all the sodium present as car-
bonate, sulphide, and hydrate. Pipette out 20 cubic centi-
meters of the clear liquid from above and titrate, as usual,
with normal hydrochloric acid, using methyl orange as indi-
cator. Each cubic centimeter of acid corresponds to 1. 55 per
cent, of Nafi.

4. Sodium Sulphide. — Pipette out 10 cubic centimeters
of the clear liquor from the supply flask, dilute to about
200 cubic centimeters, acidify with acetic acid, and titrate
with ^ normal iodine solution, using starch paste as indica-
tor. Each cubic centimeter of iodine solution used equals
.39 percent, of sodium sulphide, and is equivalent to .1 cubic
centimeter of normal acid.

5. Caustic Soda, — Pipette out 40 cubic centimeters of
the clear liquid from the supply flask into a 100-cubic-centi-
meter measuring flask, add 20 cubic centimeters of a 10-per-
cent, barium-chloride solution, and fill to the mark with
water. Thoroughly shake and allow to settle. Pipette out
60 cubic centimeters and titrate with normal hydrochloric
acid, using methyl orange as indicator. This titration
gives both sodium hydrate and sodium sulphide. To
determine the hydrate alone, multiply the number of
cubic centimeters of iodine solution used above by 20 and

§ 31 ALKALIES AND HYDROCHLORIC ACID 23

subtract the product from the number of cubic centimeters
of normal acid used here. The remainder gives the number
of cubic centimeters of normal acid used for the caustic soda,
and each cubic centimeter equals 2 per cent, of NaOH,

6. Sodium Carbonate, — Subtract the total amount of
hydrochloric acid used for the sodium hydrate and the
sodium sulphide above from the amount used for the total
alkali, and the difference gives the number of cubic centi-
meters of normal acid used for the sodium carbonate. Each
cubic centimeter of normal acid equals 2.65 per cent, of
sodium carbonate.

7. Salt, — Pipette out 10 cubic centimeters of the clear
liquid from the supply flask and titrate according to Vol-
hard's method for chlorine. All the chlorine is calculated
as salt, and each cubic centimeter of the jV normal silver
nitrate solution used equals .58 per cent, of salt.

8. Sodium Sulphate, — Pipette out 20 cubic centimeters
of the clear liquid from the supply flask and add hydro-
chloric acid in slight excess. Boil to expel carbon dioxide
and precipitate hot, as usual, with barium chloride. The
weight of barium sulphate multiplied by .3047 gives the per-
centage of sodium sulphate.

34. Lye From Extraction of Black Asli.— The follow-
ing determinations are made :

1. Specific Gravity. — Determine the specific gravity of
the warm lye by means of the Baum^ hydrometer.

2. Total Alkali, — Determine the total alkali in 2 cubic
centimeters of the lye, as under Art. 23.

3. Sodium Sulphide, — Determine the sodium sulphide in
2 cubic centimeters of the lye, as under Art. JJ3.

4. Caustic Soda. — Determine the caustic soda in 2 cubic
centimeters of the lye, as under Art. 23.

5. Sodium Carbonate. — Determine the sodium carbonate,
as under Art. 23.

6. Salt, — Determine the salt in 2 cubic centimeters by
Volhard's method, described in Quantitative Analysis,

24 ALKALIES AND HYDROCHLORIC ACID § 31

7. Sodium Sulphate, — Determine the sodium sulphate in
5 cubic centimeters, as under Art. 33«

8. Total Sulphur. — Treat 5 cubic centimeters of the lye
with an excess of bleaching powder and hydrochloric acid
(the chlorine must smell strongly). Boil off the chlorine,
filter from insoluble matter, and precipitate with barium
chloride, as usual.

9. Sodium Ferrocyanide, — Acidify 30 cubic centimeters
of the lye with hydrochloric acid and add, with constant
stirring, a strong solution of bleaching powder from a
burette, until a drop taken out shows no blue color with a
ferric-chloride solution. The ferric chloride must be free
from ferrous salts, and the end point must be quite accu-
rately reached, although a drop or two in excess does no
harm. This oxidizes the sodium ferrocyanide completely to
sodium ferricyanide. Add to the oxidized solution -^ nor-
mal copper sulphate from a burette, until a drop of the solu-
tion no longer gives a blue color with ferrous sulphate, but
shows a red color. This indicates that no more sodium
ferricyanide is present in the solution, and that the ferrous
sulphate is reducing the yellowish copper ferricyanide to the
reddish copper ferrocyanide. The first decided red color
must be taken as the end point, even if it disappears after a
time.

The copper-sulphate solution is made by dissolving
12.457 grams of crystallized copper sulphate in 1,000 cubic
centimeters of water and standardizing it against pure non-
effloresced potassium ferrocyanide.

10. Silica^ Ferric Oxide^ and Alumina, — Acidify 100 cubic
centimeters of the lye with hydrochloric acid, heat to boil-
ing, add about 1 gram of ammonium chloride, and precipitate
with ammonia. Heat until the ammonia odor is very faint,
filter, ignite, and weigh as usual.

25. Carbonated Lye. — The determinations are made as
above, but in addition the sodium bicarbonate is determined.

§ 31 ALKALIES AND HYDROCHLORIC ACID 25

Sodium Bicarbonate. — The method given in Art. 13
cannot be satisfactorily used here, for the sulphide that may
he present will interfere with the test. The following
method, however, gives good results when carefully carried
out. A standard solution of caustic soda free from carbon
dioxide is required and is best prepared by dissolving
50 grams of the best caustic soda in 1 liter of water and
adding barium chloride to precipitate all the carbon dioxide.
The solution is then standardized by acid as usual, made to
normal, and preserved as under Art. 13. For the analysis,
take 50 cubic centimeters of the carbonated lye and add
30 cubic centimeters of the caustic-soda solution, then an
excess of a lO-per-cent. barium-chloride solution, and finally
titrate with normal hydrochloric acid, using phenol-phthalein
solution as indicator. The difference between the amount
of caustic-soda solution taken and the normal acid required
gives the number of cubic centimeters of normal caustic
soda required for the bicarbonate present, and each cubic
centimeter equals .084 gram of sodium bicarbonate.

For example, if 25 cubic centimeters of normal acid is
required to titrate back, then 30 — 25 = 5 cubic centimeters
of caustic soda required for the bicarbonate present. There-
fore, .084 X 5 = .42, and .42 X 20 = 8.4 grams of sodium
bicarbonate per liter of lye.

26« Red lilquors. — The red liquor may be analyzed the
same as the crude lye, except that in the case of crude lye
all the oxidizable sulphur compounds are assumed to be sul-
phides. In the case of a red liquor, however, through oxida-
tion and other changes the sulphite and thiosulphate become
prominent and must be determined, especially when the red
liquor is used for the manufacture of caustic soda.

1. Sodium Sulphide, Sulphite, Thiosulphate, and Sulphate,
{a) Determine the total alkalinity by titrating 25 cubic centi-
meters of the liquor with normal acid, using methyl orange as
indicator. This gives sodium carbonate, sodium hydrate,
sodium sulphide, and one-half of the sodium sulphite {Na^SO^
is alkaline to methyl orange, while HNaSO^ is neutral).

26 ALKALIES AND HYDROCHLORIC ACID § 31

(d) Acidify 25 cubic centimeters of the liquor with dilute
acetic acid and titrate with -^ normal iodine solution. This
gives sodium sulphide, sodium sulphite, and sodium thio-
sulphate.

(c) Take 50 cubic centimeters of the liquor and pre-
cipitate it with an alkaline-zinc solution, make to 200 cubic
centimeters, and take 100 cubic centimeters. Acidify this
with dilute acetic acid and titrate with ^ normal iodine solu-
tion. This gives sodium sulphite and sodium thiosulphate.

(d) Take 100 cubic centimeters of the liquor and add
an excess of a 10-per-cent. barium-chloride solution to pre-
cipitate the sulphite, make up to 200 cubic centimeters,
cork tight, and allow to settle clear (or filter) ; then take
50-cubic-centimeter portions of the clear liquid for titration.
(1) Titrate a 50-cubic-centimeter portion with normal hydro-
chloric acid, using methyl orange as indicator. This gives
sodium hydrate and sodium sulphide. (2) Acidify a
second 50-cubic-centimeter portion with dilute acetic acid
and titrate with ^ normal iodine solution. This gives
sodium sulphide and sodium thiosulphate.

2. TAe Calculation.^ — ^ — ^ (2) = ^ cubic centimeters
-^ normal iodine solution corresponding to sodium sulphite.

b — c = B cubic centimeters ^ normal iodine solution
corresponding to sodium sulphide.

rf(2) — B = Ccubic centimeters ^ normal iodine solution
corresponding to sodium thiosulphate.

d (1) — -^ B = D cubic centimeters normal acid solution
corresponding to sodium hydrate.

1 — [d (1) -\- -j^ A] = £ cubic centimeters normal acid
solution corresponding to sodium carbonate.

Each cubic centimeter of ^ normal iodine solution equals
.0039 gram of Na^S, .0063 gram of Na^SO^, or .0158 gram
of Na,S,0,.

Each cubic centimeter of normal acid equals .04 gram of
NaOH, or .053 gram of Na^CO^,

21. Tank Waste. — Samples are collected in wide-mouth
glass-stoppered bottles and kept closed until analyzed. The

§ 31 ALKALIES AND HYDROCHLORIC ACID 27

determinations are made on the moist substance, as any
attempt to dry it inevitably leads to oxidation, and so to a
change of composition.

1. Alkaline Sodium Compounds. — Stir 20 grams of tank
waste thoroughly together with about 175 cubic centimeters
of warm water, let stand 1 hour to thoroughly settle, and
pour oflf the clear liquid. Pass carbon dioxide for 5 minutes,
and boil to about one-half of the original volume, to decom-
pose calcium bicarbonate and precipitate calcium carbonate.
Filter and titrate the filtrate with normal acid, using methyl
orange as indicator. Each cubic centimeter of normal acid
equals .031 gram of Na^O.

2. Total Sodium Compounds, — Heat 17.7 grams of the
waste in a porcelain dish with sulphuric acid of 50° Baum6
until the w^aste is completely decomposed, heat to drive oflf
all the free acid, add hot water, and bring into a 250-cubic-
centimeter measuring flask. Add milk of lime (made by
slaking lime, shaking up with water, pouring off one portion
to remove alkalies and then shaking up with water and filter-
ing) to remove any free acid and magnesia, fill to the mark,
let settle, and pipette off 50 cubic centimeters. To this
50 cubic centimeters add 10 cubic centimeters of a saturated
barium-hydrate solution and filter through a dry filter.
Take 50 cubic centimeters of the filtrate and precipitate all
the barium by carbon dioxide and boiling. Filter and titrate
the filtrate with normal hydrochloric acid, using methyl
orange as indicator. When the above amount of substance
is taken and allowance is made for the precipitates in the
volumes, each cubic centimeter of normal acid used equals
1 per cent, of Nafi,

FINISHED PRODUCTS

28. Soda Ash. — The determination of silica, sodium
sulphate, sodium chloride, ferric oxide and alumina, calcium
carbonate, and magnesium carbonate is carried out as under
Art. 17. In addition to these substances, it is necessary to

28 ALKALIES AND HYDROCHLORIC ACID § 31

determine in Le Blanc soda, total alkali, sodium carbonate,
caustic soda, sodium sulphide, and sodium sulphite.

1. Total Alkali, — Dissolve 3.1 grams of the soda ash in
about 150 cubic centimeters of distilled water and titrate
with normal sulphuric acid, using methyl orange as indi-
cator. Each cubic centimeter of the acid used equals 1 per
cent, of Nafi,

2. Sodium Carbonate. — Calculate from determinations
3 and 4 (below) the equivalent percentages of Nafi and
deduct the sum of these results from the percentage of Nafi
found in 1. The remainder is the alkali equivalent of the
sodium carbonate, and this remainder multiplied by 1.71 gives
the percentage of sodium carbonate in the soda ash. For
example, if 58 cubic centimeters of normal acid is used in 1,
10 cubic centimeters of ^ normal acid in 3, and 5 cubic
centimeters of silver nitrate in 4; according to 1, we have
58 per cent, of Nafi^ according to 3, .31 per cent, of Nafi
as NaOH^ and according to 4, .39 per cent, of Nafi as NaJS\
or .31 + .39 = .7 percent, of Nafi in the substance in other
forms than sodium carbonate and 58 — .7 = 57.3 per cent,
of Nafi as sodium carbonate. Then 57.3 X 1.71 = 97. 98 per
cent, of sodium carbonate in the soda ash.

3. Caustic Soda. — Dissolve 10 grams of the soda ash in
about 75 cubic centimeters of water, add an excess of a
10-per-cent. barium-chloride solution, and titrate with ^nor-
mal hydrochloric acid, using phenol phthalein as indicator.
Each cubic centimeter of acid used equals .04 per cent of
NaOH and is equivalent to .031 per cent, of Nafi,

4. Sodium Sulphide. — Dissolve 5 grams of the soda ash
in about 100 cubic centimeters of water, heat nearly to boil-
ing, and make strongly alkaline with ammonia. Titrate
with an ammoniacal silver-nitrate solution until no more
silver sulphide forms. Near the end it is advisable to filter
off a little and test to make sure of the end point.

To make the standard silver solution, dissolve 13.845 grams
pure silver in pure nitric acid, add 250 cubic centimeters of
strong ammonia water, and dilute to 1 liter. Each cubic

§ 31 ALKALIES AND HYDROCHLORIC ACID 29

centimeter of the silver solution equals . 1 per cent, of sodium
sulphide, and is equivalent to .0795 per cent, of Nafi,

5. Sodium Sulphite, — Dissolve 5 grams of the soda ash in
about 50 cubic centimeters of water, acidify with acetic
acid, and titrate with jV normal iodine solution. Each cubic
centimeter of iodine solution equals .126 per cent, of sodium
sulphite.

189. Crystal Soda. — This substance is analyzed in the
same manner as the above, except that on account of the
large amount of water of crystallization, about double
the amount must be taken for analysis.

CHANCE-CL.AU8 SULPHUH RKCOVEUY

30. Available Sulphur In Tank Waste. — In this deter-
mination the sulphide sulphur is set free by hydrochloric
acid, collected in sodium-hydrate solution, and after acidi-
fying, titrated with iodine solution. The details of the
process are as follows. Weigh out in a 500-cubic-centimeter
flask 2 grams of the tank waste, insert a two-holed rubber
stopper through one hole of which is passed a funnel tube
with a stop-cock, and through the other a tube bent to
connect by means of a tight rubber tube, a suitable absorp-
tion apparatus. The apparatus described for the determi-
nation of sulphur in iron by evolution, in Quantitative
Analysis^ is suitable for this purpose. Two of the absorption
tubes should be partially filled with sodium-hydrate solu-
tion and connected to the evolution flask. Slowly run
hydrochloric acid (1 part of acid to 1 part of water) through
the funnel tube on to the waste until the decomposition is
completed. Boil the flask, to drive out all the hydrogen
sulphide, and when the first absorption tube has become
warm on account of the steam condensed in it, open the
stop-cock of the funnel tube and allow the apparatus to cool.
Empty the absorption tubes into a 5()()-cubic-centimeter
measuring flask, fill to the mark with well-boiled water, and
take 50 cubic centimeters for titration. Dilute this to

30 ALKALIES AND HYDROCHLORIC ACID § 31

200 cubic centimeters with well-boiled water, acidify with
acetic acid, and titrate with ^ normal iodine solution.
Each cubic centimeter of the iodine solution equals
.0017 gram H^S or .0016 gram 5.

31. lilme-Klln Gases. — Determine carbon dioxide,
oxygen, and carbon monoxide as under Art. 8.

32. Gas From the Gasometer. — Determine hydrogen
sulphide and carbon dioxide together by absorbing them in
caustic-potash solution in the same manner as carbon dioxide
is determined in Art. 8.

Determine hydrogen sulphide alone by fitting a flask of
exactly known content (about 500 cubic centimeters) with a
two-holed rubber stopper, through one hole of which passes
a funnel tube with a glass stop-cock; the stem of the funnel
tube should end just below the stopper. Through the other
hole in the stopper passes a tube, which leads to the bottom
of the flask and is fitted with a stop-cock. For making the
determination, allow gas from the gasometer to pass through
the apparatus until the air is completely displaced, close both
stop-cocks, disconnect from the gasometer, and empty the
gas from the tubes outside of the stop-cocks. Run in through
the funnel tube about 25 cubic centimeters of a normal
sodium-hydrate solution and shake thoroughly until all
the gas is absorbed. Wash out into a 250-cubic-centimeter
flask with air-free water and make to the mark on the flask.
Take 50 cubic centimeters, dilute to about 250 cubic centi-
meters with air-free water, acidify with acetic acid, and
titrate with standard iodine solution. The standard iodine
solution should contain 11.43 grams of iodine per liter, when
each cubic centimeter equals 1 cubic centimeter of hydrogen-
sulphide gas at O'' C. and 760 millimeters of mercury pres-
sure.

To reduce the gas employed to normal conditions use the
formula given for this purpose in Quantitative Analysis.

If necessary to calibrate the flask, it can be done with suffi-
cient accuracy by weighing it empty, then filling with water
to the stop-cocks, and weighing again. The difference

§ 31 ALKALIES AND HYDROCHLORIC ACID 31

between the two weights gives the weight of water in the
flask, and, therefore, the volume in cubic centimeters. If
greater accuracy is desired, the temperature of the water
may be taken and the expansion of the water above 4° C.
allowed for. Furthermore, the volume of air in the flask at
its first weighing is approximately given by the weight of
water ; the weight of the air can be deducted from the weight
of the flask plus air, thus giving the weight of the empty
flask. For example, the flask plus air weighs 300 grams,
the flask plus water at 18^ C. weighs 795 grams; then
795 — 300 = 495 grams of water at 18° C, which equals,
approximately, 495 cubic centimeters as the capacity of the
flask.

Correcting, 1 liter of air under standard conditions weighs
1.293 grams; and if the barometer stands at 750 millimeters
of mercury pressure, the weight of 495 cubic centimeters of
air can be calculated (see Quantitative Analysis). For

'V = jT— ^— j = 458 cubic centimeters at standard

conditions = .458 liter. Therefore, 1.293 X .458 = .6 gram,
the weight of air in the flask. The real weight of the flask
is, therefore, less by this amount than the apparent weight
and the weight of water becomes 495. 6 grams. But 1 gram
of water at 18° C. equals 1.001373 cubic centimeters, and
therefore the corrected volume of the flask is 496.3 cubic
centimeters.

33» Waste Gas From Claus Kiln. — The important sub-
stances to determine in this gas are sulphur dioxide and
hydrogen sulphide. These are best determined by conduct-
ing 5 liters of the gas through a suitable absorption appar-
atus containing caustic-soda solution. The gases are
absorbed, giving sodium sulphide and sodium sulphite.
The caustic solution is then made to 250 cubic centimeters
with air-free water and 50 cubic centimeters taken, acidified
with acetic acid, and titrated with ^^ normal iodine. This
gives both the hydrogen sulphide and sulphur dioxide.
100 cubic centimeters of the original solution is then taken.

33 ALKALIES AND HYDROCHLORIC ACID § 31

the sulphide precipitated with an alkaline* zinc solution,
one-half filtered oflf, acidified with acetic acid, and titrated
with ^^ normal iodine solution; this g^ves the sulphur
dioxide. 1 cubic centimeter of ^ normal iodine solution
equals .0017 gram of H^S or .0032 gram of S(?„ and equals
1.12 cubic centimeters of either gas at 0" C. and 760 milli-
meters of mercury pressure.

SODIUM BICARBONATE

34. The crude materials for sodium bicarbonate manu-
facture are the soda crystals from Le Blanc soda or the
ammonia-soda ash, and lime-kiln gas. For the analysis of
these substances, see Arts. 8, 17, and 29.

FINISHED PRODUCT

35. Hoiliuiii Bicarbonate. — Analyze the same as soda
ash, Art. 17. The daily tests consist in the determination
of total alkali, sodium carbonate, sodium bicarbonate, and
sodium chloride.

CAUSTIC SODA

CRUDE MATERIAr.S

36. The crude materials for the manufacture of caustic
soda differ, depending on whether the substance is made at
a Le Blanc or at an ammonia-soda works. The methods
for all of them, however, will be described, and the student
can select those that apply to the work that he is doing.

1. Red Liquor — Analyze as under Art. 26.

2. Soda Ash. — Analyze as under Art. 17 or !88«

3. Milk of Livie, — Analyze as under Art. 15.

31 ALKALIES AND HYDROCHLORIC ACID

IXTKRMEDIATE PUODUCrrS

37. While some of the following may be very properly
considered as finished products, or otherwise classified, for
the sake of simplicity they are given under this head.

38. Caustic I^lquor. — The following determinations are
made:

1. Specific Gravity. — The specific gravity is taken of the
liquor at different stages of the evaporation, and although
other substances affect the results, a fair idea of the run of
the liquor can be obtained by this determination alone.
.Table III gives the percentage of caustic soda corresponding
to the different specific gravities at 15"" C.

TABIiE

ITT

Specific
Gravity

Grams of

NaOH

Per Liter

Sp)ecific
Gravity

1
Grams of j ^

XaOH ,,P

Per Liter

•ecific
avity

Grams of

NaOH

Per Liter

1.007

1. 142

144 I,

320

381

1. 014

1. 152

156 ; I.

332

399

1.022

1. 162

167

345

420

1.029

1. 171

177 I.

357

441

1.036

1. 180

188 1 I.

370

462

1.045

1. 190

200 I.

Z^^

483

1.052

1.200

212 I.

397

506

1.060

1. 210

225 I.

•

410

528

1.067

1.220

239 I

424

553

1.075

1. 231

253 I

438

575

1.083

1. 241

266 I

•453

602

1. 091

1.252

283

.468

629

1. 100

1.263

299 I

.483

658

1. 108

104

1.274

316 I

.498

691

1. 116

112

1.285

332 I

•514

721

1. 125

123

1.297

348 I

•530

750

1. 134

134

1.308

364

34 ALKALIES AND HYDROCHLORIC ACID § 31

2. Total Alkali and Sodium Carbonate, — These two val-
ues are determined according to Art. S3.

3. Salt. — In caustic from ammonia soda, it is frequently
necessary to determine the amount of salt. Proceed accord-
ing to Volhard's method, described in Quantitative Analysis,

4. It is only necessary to determine sulphur compounds
when the caustic is made from red liquor or crude Le Blanc
soda. Sodium sulphate is sometimes determined in liquor
from ammonia soda. Make the determinations according to
Art. 26.

39. Fished Salts. — For analysis dissolve 26 grams of
the salts in 600 cubic centimeters of water.

1. Total Alkali, — Titrate 26 cubic centimeters, as usual,
with normal acid, using methyl orange as indicator.

2. Salt, — Titrate 25 cubic centimeters with silver nitrate
by Volhard's method, described in Quantitative Analysis.

3. Sodium Sulphate, — Determine in 26 cubic centimeters,
by acidifying with hydrochloric acid and precipitating hot
with barium chloride, as usual.

4. Oxidizable Sulphur Compounds. — Treat 26 cubic centi-
meters of the solution with bromine water until it is colored,
acidify with hydrochloric acid, boil off the excess of bro-
mine, and precipitate as sulphate with barium chloride as
usual. The difference between the amount of sulphate
found here and that found above gives the oxidizable sul-
phur. This determination is, of course, unnecessary when
the caustic is made from ammonia soda.

40. Caustic Bottoms. — This sample sometimes comes
to the laboratory in fairly large lumps in a stoppered bottle
that has the stopper covered with sealing wax. This wax
should not be broken until the sample is wanted for analy-
sis. Then several pieces are taken, wrapped quickly in sev-
eral thicknesses of heavy brown paper, and crushed on an
anvil by means of a hammer; 20 grams are then weighed
oflf and dissolved in water. It is necessary to work quickly

§ 31 ALKALIES AND HYDROCHLORIC ACID 35

until the caustic is weighed, to prevent its absorbing water
from the air.

1. Insoluble.' — When the above 20 grams are dissolved,
filter, and wash thoroughly. Collect the filtrate and wash-
ings in a 500-cubic-centimeter measuring flask, make to the
mark, and save. The filter and contents are ignited and
weighed.

2. Total Alkali, — Take 50 cubic centimeters of the above
filtrate, add a little lacmoid for an indicator, and add nor-
mal acid to more than neutralize. Hoat to boiling, to expel
the carbon dioxide, and titrate back with normal alkali.
The difference between the acid and alkali used gives the
acid required for neutralizing the total alkali. Each cubic
centimeter of normal acid equals .031 gram of Na^O.

3. Sodium carbonate is determined according to Art. 23.

4. Salt is determined according to Art. 23.

41. Caustic Mud. — The determinations are as follows :

1. Total Alkali, — Extract 25 grams of the sample by
shaking it with several small portions of hot water, finally
filter, wash, and unite the filtrates and washings, pass carbon
dioxide for 10 minutes, boil to decompose bicarbonates,
refilter, if necessary, and titrate with normal acid, using
methyl orange as indicator. Each cubic centimeter equals
.031 gram of Na^O.

2. Caustic Lime, — Shake about 25 grams of the waste
with a little water and titrate with normal acid and phenol
phthalein. The sodium above was present as hydrate and
carbonate, but a fair average will be reached if we deduct
one-half of the number of cubic centimeters of acid required
for total alkali, from the amount taken above, and call the
remainder of the acid used by the waste, caustic lime.
Each cubic centimeter of acid equals .037 gram of Ca(OH)^,

3. Calcium Carbonate. — Titrate 1 gram of the sample
with normal hydrochloric acid, using methyl orange as indi-
cator, and deduct the acid required for caustic lime. Each
cubic centimeter of acid equals .05 gram of CaCO^

36 ALKALIES AND HYDROCHLORIC ACID § 31

FINISHED PRODUCTS

42, Caustic Soda. — The method for preparing the
sample for analysis given under Art. 40 can be used to
advantage here. For analysis weigh out 50 grams, dissolve
in water, and make to 1,000 cubic centimeters.

1. Total Alkali, — Titrate as usual, using normal hydro-
chloric acid and methyl orange.

2. Caustic soda is determined as under Art. 23.

3. Sodium carbonate is determined as under Art. 23.

4. Salt is determined as under Art. 23.

5. Sodium sulphate is determined as under Art. 39.

6. Other constituents are determined as under Art. 17.

HYDROCHIiORIC ACID

RAW MATERIALS AND INTERMEDIATE PRODUCTS

43. Hydrochloric acid is almost without exception ob-
tained from salt by the action of sulphuric acid. For its
crude materials and intermediate products, see under the
heading ** Salt Cake."

The absorption of the gas in the bombonnes and towers
is watched by means of specific-gravity tests. These are
best made by arranging a cylinder and hydrometer in such
a way that a portion of the acid is being continuously
collected in the cylinder in which the hydrometer floats. By
this means it is possible to see the specific gravity at a
glance, and the delay and trouble of collecting the sample is
avoided.

Table IV gives the specific gravity and composition of
solutions of hydrochloric acid at 15° C.

44. "Waste Gases. — The gas that escapes from the
absorption towers must not contain much hydrochloric acid,
for it is injurious to vegetation. The sample is taken by

§ 31 ALKALIES AND HYDROCHLORIC ACID 37

inserting a glass tube to the center of the chimney through
which the gas passes to the outside air. To the outer end
of the tube is attached a double-acting rubber suction bulb,
and this, in turn, is connected to an absorption apparatus

TABLE IV

Specific
Gravity

Per Cent.
HCl

Grams

HCl

per Liter

Specific
Gravity

Per Cent.
HCl

Grams

HCl

per Liter

I.OOO

.16

1.6

1.1150

22.86

, 255

1.005

1. 15

12.0

1. 1200

23.82

267

1. 010

2.14

22.0

1. 1250

24.78

278

1. 015

3.12

32.0

1. 1300

25-75

291

1.020

4.13

42.0

1-1350

26.^70

303

1.025

5-15

53.0

1. 1400

27.66

315

1.030

6-15

64.0

1. 1425

28.14

322

1.035

7-15

74.0 ,

1. 1450

28.81

328

1.040

8.16

85-0 \

1. 1500

29-57

340

1-045

9.16

96.0 1

1. 1520

29.95

345

1.050

10.17

107.0

I 1550

30.55

353

1.055

II. 18

118.0

1. 1600

31-52

366

1.060 '

12.19

129.0

1. 1630

32.10

373

1.065

13.19

141. 0

1. 1650

32.49

379

1.070

14.17

152.0 1

1. 1700

33- 46

392

1.075

15.16

163.0 .

1.1710

33.65

394

1.080

16.15

174.0

1. 1750

34.42

404

1.085

17.13

186.0

1. 1800

35-39

418

1.090

18.11

197.0

1. 1850

36.31

430

1.095

19.06

209.0

1. 1900

37-23

443

1. 100

20.01

220.0

K1950

l^^'i^^

456

1. 105

20.97

232.0

1.2000

39-11

469

I. no

21.92

243.0

similar to that mentioned in Art. 30, The absorption
apparatus is fitted with large test tubes, or small flasks, so
that two pieces will hold 150 or 200 cubic centimeters of
water. It is then filled with water and is connected in

38 ALKALIES AND HYDROCHLORIC ACID g 31

position. The bulb is then compressed a sufficient number
of times to force the desired amount of chimney gas through
the absorption apparatus. By careful work the amount of
gas used can be quite accurately estimated by this method;
if greater accuracy is wished, the gas after passing through
the absorbing apparatus may be run into a gasometer and
measured. The liquid from the absorption apparatus is
washed into a flask and titrated by
Volhard's method, which is de-
scribed in Quantitative Analysis.

Another very simple and very
effective form of absorption appara-
tus that can be used has been
recommended by the English
alkali inspectors ; it is shown in
Fig. 3. The gas enters at a,
passes out through the holes at the
lower end of the tube, and passes
up through a number of thin ends
cut from a small rubber tube,
which breaks the gas into fine
bubbles, then out through the holes,
in the direction of the arrows,
into the bottle, and Anally escapes
through the tube b. This tube is
filled below with pieces of rubber
tube and above with glass wool.
p,g J By moistening the contents of b

with water and adding a little indi-
cator, as methyl orange, any failure on the part of the
apparatus to absorb the acid is shown in b by the change in
the indicator.

FINISHED PRODUCT

45. Hydrocbloiic Add. — The analysis of hydrochloric
acid varies according to the purpose for which the acid is to
be used. For many purposes a simple determination of the
specific gravity is sufficient, while for other purposes a more

g 31 ALKALIES AND HYDROCHLORIC ACID 39

extended examination is necessary. In the following, the
methods of analysis are given for all cases, except the so-
called chemically pure acid, the examination of which is
practically neverrequired in the ordinary chemical works.

1. Sulphuric Acid, — Take 50 cubic centimeters of the acid
to be tested, almost neutralize with pure sodium carbonate,
heat to boiling, and precipitate with barium chloride, as
usual. Each gram of barium sulphate found corresponds
to .34335 gram of SO^.

Another method, which gives quite accurate results and,
on account of its rapidity, is very suitable where several
determinations must be made each day, is as follows : Pre-
pare a glass tube 6 millimeters broad and 250 millimeters
long closed at the lower end, while the upper end expands
into a tube 15 millimeters broad. Provide a rubber stopper
for the broad tube. By mixing acids of known composition
make a series of acids containing from .2 or .6 up to 3 per
cent, of sulphuric acid. Take 10 cubic centimeters of the
first of these acids, heat to boiling, pour into the above tube,
nearly neutralize with ammonia, and precipitate with 5 cubic
centimeters of a boiling hot, saturated, barium-chloride
solution. Insert the rubber stopper, place in a centrifugal
machine, and whirl for 5 minutes. Mark the height of the
precipitate, empty, and repeat with the next stronger sam-
ple. In this way graduate the tube and use it for the deter-
mination in the same way, using 10 cubic centimeters of
the sample, instead of the known solution, and reading off
the percentage of sulphuric acid on the tube.

2. Sulphurous Acid, — Add bromine to 50 cubic centi-
meters of the acid to color it and boil until color disappears.
Proceed as for sulphuric acid. For rapid work, use 10 cubic
centimeters of the sample and use the rapid method given
above. In either case, deduct the barium sulphate found
above from the total and each gram of barium sulphate in
excess corresponds to .27468 gram of SO^,

3. Arsenic, — The detection and determination of arsenic
in hydrochloric acid that is to be used in the preparation of

40 ALKALIES AND HYDROCHLORIC ACID § 31

foodstuffs IS very important. A very large number of
methods for both its qualitative and quantitative determi-
nation have been proposed and are in use. The following,
however, seem to be the most convenient and exact.

{a) Qualitative Tests. — Take 10 cubic centimeters of the
sample in a test tube, dilute with 10 cubic centimeters of
distilled water, carefully pour on the top of the acid 5 cubic
centimeters of a freshly prepared hydrogen-sulphide solution,
and allow to stand for 1 hour. Prepare a second tube in
exactly the same manner and allow to stand for 1 hour in a
water bath at from 70° to 80° C. If no precipitate, or yellow
ring, appears between the two layers in either case, arsenic
is absent. By this method the presence of ^ milligram of
arsenic in the 10 cubic centimeters of acid can be detected.

For the most accurate detection of arsenic take 5 liters of
the acid, add about \ gram of potassium chlorate, to prevent
the arsenic volatilizing as AsCl^ during evaporation, and
dilute with water until the specific gravity does not exceed
1.1. Evaporate to dryness in a well-enameled porcelain
evaporator, take up the residue in a little water, and test the
solution in a Marsh apparatus, which is described in Quali-
tative Analysis.

{b) Quantitative Determination. — When very small
amounts of arsenic are to be determined, take 5 liters of the
acid, and concentrate to small bulk as above, using potas-
sium chlorate to prevent loss of arsenic by volatilization,
then proceed as follows: If fairly large amounts are known
to be present or are shown by the qualitative test, take
50 cubic centimeters, partly neutralize with sodium carbonate,
dilute to 150 cubic centimeters, and precipitate as sulphide,
following the directions given in Quantitative Analysis.
Remember here that the arsenic may be present as arsenic
acid and that, under those circumstances, heat and consider-
able time (from 12 to 20 hours) are necessary to completely
precipitate all the arsenic.

4. Selenium. — Test with stannous chloride as described
in Qualitative Analysis.

§ 31 ALKALIES AND HYDROCHLORIC ACID 41

5. Hydrochloric Acid, — Take 10 cubic centimeters of
the sample in an accurate pipette, dilute to 250 cubic centi-
meters, and take 25 cubic centimeters for titration. Titrate
with normal caustic-soda solution, using methyl orange as
indicator. Deduct the amount of caustic corresponding to
the SO^ already found from the total and the rest corre-
sponds to HCl. Each cubic centimeter of alkali equals
.0365 gram of HCL

For example, if 10 cubic centimeters of normal alkali is
required for 1 cubic centimeter of the sample and .004 gram
of 5(9, has been found in the previous determination, then
. 1 cubic centimeter of the alkali was used by the sulphuric
acid, and the amount used by the hydrochloric acid is
9.9 cubic centimeters, which equals .36135 gram of HCl in
1 cubic centimeter of the sample, or 361.35 grams per liter.

It is customary to repx)rt results of this kind in grams per
liter; but if the percentage is wanted, determine the specific
gravity and divide the grams per liter by 10 times the
specific gravity, the result will be the percentage of HCL

When the amount of hydrochloric acid alone is to be
determined in a sample, it is simpler to titrate 10 cubic cen-
timeters of the diluted sample with ^ normal silver nitrate,
using Volhard's method, which is described in Quantitative
Analysis. Each cubic centimeter of the silver-nitrate solu-
tion equals .00365 gram of HCL

CHIiORINB, BliEACHING COMPOUNDS, CHLORATES

CRUBB MATERIALS

46, Mansranese Ore, — The ordinary determinations are
as follows :

1. Moisture, — Spread 2 grams of the finely powdered ore
thinly on a watch glass and dry at lOO"* or llO"* C. until the
weight remains constant.

42 ALKALIES AND HYDROCHLORIC ACID § 31

2. Available Oxygen. — For this determination are needed
a \ normal potassium-permanganate solution and a ferrous-
sulphate solution made
by dissolving 100 grams
of ferrous sulphate and
100 grams of sulphuric
acid in 1 liter of water.
For the determination,
weigh out 1.0875 grams
of the dried ore (prefer-
ably that used for the
moisture determination)
P'o- 4 into a 200-cubic-centi-

meter flask provided with a tube leading to the bottom of a
second flask containing sodium-bicarbonate solution. The
arrangement of the flasks is shown in Fig. 4. Measure
exactly 75 cubic centimeters of the ferrous sulphate into the
flask with the manganese ore, insert the stopper with the
tube leading into the sodium-bicarbonate solution, and heat
until a dark-colored residue is no more apparent. Allow
the solution to cool, wash into a 500-cubic-centimeter beaker,
dilute to about 200 cubic centimeters, and titrate with \ nor-
mal potassium-permanganate solution until the color stays
permanent for about \ minute.

The ferrous-sulphate solution must be standardized each
day by measuring out 75 cubic centimeters, using the same
pipette as above, and titrating it with the \ normal potas-
sium-permanganate solution.

The difference between the amount of potassium-perman-
ganate solution used to titrate the ferrous-sulphate solution
and that used with the ore gives the available oxygen, or
rather the manganese present in the ore as MnO^, If the
above amount of ore is weighed out, each cubic centimeter
of \ normal potassium-permanganate solution corresponds
to 2 per cent, of MnO^,

Another very exact and rapid method that can be used
direct, or as a check on the above method, is given in Quan-
titative Analysis^ under the description of the nitrometer.

g 31 ALKALIES AND HYDROCHLORIC ACID 43

3. Carbon Dioxide, — Determine according to the absorp-
tion method given in Quantitative Analysis.

4. Acid Necessary to Decompose Ore, — Bring 1 gram of
the ore into a flask containing 10 cubic centimeters of the
hydrochloric acid being used in the chlorine manufacture and
whose titration strength has been previously determined.
Insert a stopper, with a return condenser, in the flask and
heat until the ore is dissolved. Allow to cool and titrate
with normal caustic-soda solution until the brown flakes of
iron hydrate no longer dissolve by shaking. The differ-

" ence between the caustic soda used here and that required
for the titration of 10 cubic centimeters of the original acid
gives the acid used in decomposing the ore.

47, lilmestone. — Analyze according to Art. 3.

48, Qaicklime. — Analyze according to Art. 4,*

49, Slaked liime. — ^Water, carbon dioxide, and calcium
hydrate are usually determined.

1. Water, — Weigh out from a well-closed weighing
tube 1 gram of the sample into a weighed platinum crucible
and heat, at first gradually and then to the strongest tem-
perature of the blast lamp ; cool ; and weigh. The loss of
weight equals carbon dioxide and water.

2. Carbon Dioxide. — Determine according to the absorp-
tion method given in Quantitative Analysis and deduct the
result from the carbon dioxide and water previously deter-
mined.

3. Milk of Lime. — See Art. 16.

INTERMEDIATE PRODUCTS

60, Free Acid In Still Ijlquor. — Titrate 25 cubic centi-
meters of the still liquor with normal sodium-hydrate
solution until the brown flakes of ferric hydrate no longer
dissolve by thorough shaking. Each cubic centimeter of
caustic-soda solution used equals .0365 gram of free hydro-
chloric acid.

44 ALKALIES AND HYDROCHLORIC ACID § 31

51. Calcium Chloride In Clear lilqiior. — Acidify
25 cubic centimeters of the clear liquor with acetic acid, add
ammonium oxalate in excess, allow to stand 3 hours to insure
complete precipitation of the calcium oxalate, and filter on
an asbestos filter, using a Gooch crucible. Bring the cru-
cible containing the precipitate of calcium oxalate into a
300-cubic-centimeter beaker, add 100 cubic centimeters of
distilled water and 10 cubic centimeters of concentrated sul-
phuric acid. (Use care in adding the acid, that the contents
of the beaker do not spatter out.) Now titrate the oxalic
acid obtained from the above operations with ^^ normal
potassium-permanganate solution. Each cubic centimeter of
the ^^y normal potassium-permanganate solution is equal to
.0028 gram of calcium oxide, or .00555 gram of calcium
chloride.

62. Weldon Mud. — The following determinations are
required :

1. Manganese Dioxide. — See Art. 46,

2. Total Manganese. — Weigh out 10 grams of the mud,
acidify with concentrated hydrochloric acid, boil to drive off
all the chlorine, and then neutralize the excess of acid with
precrpitated chalk. Acidify with acetic acid, add bromine,
heat, and continue the addition until the solution retains the
odor of bromine. Add alcohol slowly until the red color
disappears and filter on a Gooch filter. Test the filtrate, to
see if it turns brown, with the addition of a drop of bromine
water ; if so, precipitate the rest of the manganese and add
it to the precipitate already obtained. All the manganese
is now on the filter as manganese dioxide. Introduce filter
and all into a flask and proceed to determine the manganese
dioxide according to Art. 46,

3. Total Base, — This indicates the base present that neu-
tralizes the hydrochloric acid without producing chlorine.
Dilute 25 cubic centimeters of normal oxalic-acid solution to
about 100 cubic centimeters, warm to 75° C, and add
10 grams of the mud. Shake until the precipitate is pure
white, dilute to 202 cubic centimeters, filter through a

g 31 ALKALIES AND HYDROCHLORIC ACID 45

dry filter, take 100 cubic centimeters of the filtrate and
titrate back with normal caustic-soda solution. (The extra
2 cubic centimeters is to allow for the precipitate.) If we
call the caustic-soda solution used x, the oxalic acid used is
25 — %x. Of this, part is used to neutralize the base, and
part to reduce the manganese dioxide to manganese monox-
ide and then neutralize that. We have just found the
amount of manganese dioxide in 10 grama of the mud and
can calculate its equivalent in oxalic acid from the equation
MnO^ + %{COOH)^ = Mn(COX + ^CO, + Zff,0
Calling this amount of oxalic acid expressed in cubic cen*
timeters of normal solution _>', then the amount of normal
acid used by the base is 25 — (Sj* +/) = -- Since the base
consists of a mixture of lime, magnesia, manganese hydrate,
and iron hydrate, it is customary to report the result here
in cubic centimeters of oxaiic acid used. .

53. Gas From tjulptiate Pan. — The hydrochloric-actd
gas from the " pan " must be mixed with the proper amount
of air as it goes to the
"decomposer," and this
mixture is controlled by
analysis. The analysis is
carried out by sucking the
gas, by means of an aspi-
rator, through a standard
solution of caustic soda
containing methyl orange.
The instant the color
changes, the fiow of the

gas is stopped and the ,

volume of gas in the aspi-
rator is determined by ^^^
measuring the amount of

water that has run out of the aspirator. A suitable piece of
apparatus for this determination is shown in Fig. 5. The
lower end of the tube leading into the absorption bottle is
blown out and arranged with a number of small holes to

46 ALKALIES AND HYDROCHLORIC ACID § 31

break up the gas into small bubbles and so assist the absorp-
tion.

By using the same amount of normal alkali each time, the
amount of hydrochloric acid absorbed is constant; and by
measuring the air carried through, the composition of the
mixed gas can be easily calculated. As, for example, if we
use 100 cubic centimeters of normal alkali that is equal to
3.65 grams hydrochloric acid, which is equal to 2.24 liters of
hydrochloric-acid gas under 0° C, and 760 millimeters of
mercury pressure. If the gas collected measures 3 liters
after correcting for temperature and pressure, then the
total gas used is 5.24 liters, of which 57.3 volume per cent,
is air and the remainder hydrochloric acid.

64, Gas rrom Decomposer. — Arrange three absorp-
tion bottles, similar to that shown in Fig. 5, in a series as
close to the decomposer as possible, and divide 250 cubic
centimeters of caustic soda of 1.075 sp. gr. between the
three bottles. The aspirator is so regulated that it con-
tinues during the working off of a pan charge. Five liters
of the gas are sucked through the absorption bottles, then
the contents of all three flasks are united and diluted to
exactly 500 cubic centimeters.

{a) Pipette off 100 cubic centimeters of the above solu-
tion, add 25 cubic centimeters of standard ferrous-sulphate
solution, and proceed as for available oxygen in Art. 46,
titrating at the end with ^ normal potassium-permanganate
solution. Deducting the amount of potassium perman-
ganate required here from the amount required for 25 cubic
centimeters of the ferrous-sulphate solution gives the amount
of the permanganate equivalent to the chlorine in 1 liter of
the gas. The number of cubic centimeters of i normal
potassium permanganate times .01775 equals the number of
grams of chlorine per liter of gas.

(d) Pipette off 25 cubic centimeters of the original solu-
tion and add somewhat of an excess of sodium-sulphite
solution (approximately the amount of sodium sulphite
needed can be estimated from the preceding determination).

§ 31 ALKALIES AND HYDROCHLORIC ACID 47

Add sulphuric acid until the solution is acid, when it should
smell strongly of sulphur dioxide, thus showing that more
sulphur dioxide is present than is needed to reduce the
sodium hypochlorite to sodium chloride. Heat to boiling,
cool, and, if necessary, add potassium -permanganate solu-
tion until the color fades out very slowly. Titrate with
fy normal silver-nitrate solution, using the Volhard method.
If the number of cubic centimeters of J normal potassium-
permanganate solution required for the chlorine under (a) is
called -x, and the number of cubic centimeters of -j^ normal

silver-nitrate solution, y, — equals the percentage decom-
position of the hydrochloric acid.

5ff. Bleacbing-Powder Chambers. — Whenever It is
necessary to open the chamber in which bleaching powder
is being made, the gas in the chamber
must be tested in some way, in order
that too much chlorine_ will not be
allowed to escape into the surrounding
atmosphere. A very simple apparatus, .
and the one in most common use for this
purpose, is shown in Fig. 6. The cylin-
der */ contains 35 cubic centimeters of a
solution made as follows: .49S gram of
arsenic trioxide is dissolved in sodium-
carbonate solution and neutralized by
sulphuric acid; then 35 grams of potas-
sium iodide, 5 grams of precipitated
chalk, and from 8 to 10 drops of ammo-
nium-hydrate solution are added, and
the whole made up to 1 liter with distilled
water. A little starch paste is added to
each 25 cubic centimeters just before
it is used. The cylinder d is fitted with ''"^' "

a two-holed rubber stopper c; through one hole passes the
tube e, which is drawn out at the lower end to a hole about
the size of a knitting needle; through the other hole passes

48 ALKALIES AND HYDROCHLORIC ACID § 31

a glass tube /, the lower end of which projects a short
distance into the cylinder, while to the upper end is attached
the rubber bulb a of about 100 cubic centimeters capacity.
The tube f is also provided with the small hole b. To
test the gas in a bleach chamber, the tube e is inserted
through an opening in the chamber about 2 feet above the
floor. The bulb a is then compressed, b is closed by the
finger, and a allowed to expand. By doing this the gas
from the chamber is drawn through the test solution in d.
By counting the number of bulbs full of the gas necessary
to color the test solution by separated iodine, the chlorine
in the gas can be calculated ; for 25 cubic centimeters of the
above solution is equivalent to 9.135 milligrams of chlorine.
That is, if it takes 10 bulbs full to bring a color, then the
gas contains 9.135 milligrams of chlorine per liter of the gas.

66, Bleach Liquors. — When liquid bleach is made
direct from the base, or carbonate, and chlorine, the manu-
facture requires a careful attention to the course of the
absorption.

1. Available chlorine IS determined in 5 cubic centimeters
of the liquid by Penot's method, which is described in
Quantitative Analysis^ under ** Bleaching Powder."

2. Chlorides. — Take the solution from the determination
of available chlorine and which now contains arsenates,
nearly neutralize with nitric acid, but still leave a slight
excess of alkali, and titrate with ^ normal silver-nitrate
solution. The formation of the red silver arsenate when
the chlorine is all precipitated shows the end point.

3. Chlorates. — Bring 5 cubic centimeters of the bleach
solution into a flask arranged as shown in Fig. 4; add
50 cubic centimeters of the solution of ferrous sulphate
described in Art. 46, and the strength of which against
J normal potassium-permanganate solution is known, boil
the mixed solution, and after allowing to cool, titrate back
with potassium-permanganate solution. If the number of
cubic centimeters of \ normal potassium-permanganate

§ 31 ALKALIES AND HYDROCHLORIC ACID 49

solution used for 50 cubic centimeters of the original ferrous-
sulphate solution is called a^ and the number of cubic centi-
meters of ^ normal potassium-permanganate solution used
by 50 cubic centimeters of the ferrous-sulphate solution after
oxidation with the bleach liquor b, then a — b gives oxidizing
equivalent of the bleach liquors in terms of \ normal potas-
sium-permanganate solution. The oxidizing action is due
to the available chlorine and the chlorates. The available
chlorine has been determined, and 5 cubic centimeters of
the -jV normal arsenite solution is equivalent to 1 cubic
centimeter of the ^ normal potassium -permanganate solu-
tion. If the number of cubic centimeters of ^^ arsenite

solution used for available chlorine is called r, then (^ — ^) — -r

equals the number of cubic centimeters of \ normal potas-
sium-permanganate solution equivalent to the chlorate in
the solution. Each cubic centimeter of ^ normal potassium-
permanganate solution is equivalent to .01021 gram potas-
sium chlorate, .00888 gram of sodium chlorate, or .00862 gram
of calcium chlorate. This gives the amount of the chlorate
in 5 cubic centimeters of the solution, which result multi-
plied by 200 gives the number of grams per liter.

4. Caustic Alkali, — Take 10 cubic centimeters of bleach
liquor and dilute with 150 cubic centimeters of distilled water,
add a few drops of a phenol-phthalein solution, and titrate
with a normal acid solution until the red color disappears.
Adda few more drops of the indicator, and if the color again
disappears after about 5 seconds shaking, the result is taken
as equivalent to the caustic alkali present.

5. Carbonates, — Take 10 cubic centimeters of the bleach
liquor and add ammonia (in a well-covered beaker to avoid
loss by the gas evolved) until the evolution of nitrogen
ceases and the liquid smells of ammonia. Then heat until
the ammonia odor disappears, dilute to 150 cubic centimeters,
and titrate with normal acid, using methyl orange as indi-
cator. The difference between this result and that for the
caustic gives the carbonate in the solution.

50 ALKALIES AND HYDROCHLORIC ACID § 31

57. Chlorates, — The methods of control here are prac-
tically the same as those described for bleach liquors under
Art. 56. The usual determinations made are chlorates,
chlorides, . and sometimes available chlorine (chlorine and
hypochlorites). The chlorate is reported as potassium
chlorate, and for calculating the amount of potassium chlo-
ride necessary to convert the calcium chlorate into potassium
chlorate, we can multiply the number of cubic centimeters
of ^ normal potassium permanganate used by 3. 105. That

is, in Art. 56 I (« — ^) — - I X 3. 105 = number of grams of

potassium chloride required per liter of the chlorate liquor
to convert the calcium chlorate into potassium chlorate.

FINISHED PRODUCTS

58. Bleaching Pcwder. — The only determination that
it is necessary to make with bleaching powder is the deter-
mination of the available chlorine. A large number of
methods have been proposed for this determination, but the
only one of importance for this country is the Penot method,
which is described in Quantitative Analysis, A somewhat
similar method, using a hydrochloric-acid solution of arsenic
trioxide, was introduced into France in 1835 by Gay-Lussac
and is still largely used in that country. It is far inferior,
however, to the Penot method.

59. Bleach. litquors. — Analyze as g^ven under the
methods for factory control, in Art. 56.

60. Potassium Chlorate. — Potassium chlorate as a fin-
ished product is so nearly chemically pure that seldom more
than a qualitative analysis is necessary, or at most a quanti-
tative determination of the chloride present. »

1. Potassium Chloride. — Dissolve 50 grams of the sample
in as little distilled water as possible, precipitate the chlorine
with silver nitrate, shake to collect the precipitate together;
filter on a Gooch filter, wash thoroughly, dry at 125° C, and

§ 31 ALKALIES AND HYDROCHLORIC ACID 51

weigh. Each gram of silver chloride is equivalent to .5192

- ^ . ,1 -J weight ofAgClx .5192 X 100
gram of potassium chloride, or — — ^^

= percentage of KCl in substance.

2. Qualitative Tests. — The solution should be water
white, free from sediment, and should not color or precipi-
tate by the addition of ammonium sulphide or carbonate.

61» Electrolysis. — The analyses required in the control
of electrolytic processes for the preparation of alkali, chlo-
rine, and potassium chlorate are so similar to those already
treated that no more than a reference to them is required.

1. Brine, — Analyze according to Art. 1.

2. Caustic liquor may contain sodium chloride, sodium
hypochlorite (possibly sodium chlorate), sodium hydrate,
and more or less sodium carbonate. Analyze according to
Art. 56.

3. Bleaching Powder Chambers, — See Art. 55.

4. Bleach Liquor. — See Art. 56.

5. Potassium-chlorate liquor may contain potassium chlo-
ride, potassium hypochlorite, potassium chlorate, and hypo-
chlorous acid. Analyze according to Art. 56.

MANUFACTURE OF IRON

ENTRODUCTOBY

!• Iron is very widely distributed in nature and its com-
pounds are abundant. Probably no portion of the earth's
crust is free from it, yet it occurs native only in very small
quantities, and the iron thus found is probably of meteoric
origin and is always alloyed to a greater or less extent with
other metals, as nickel, cobalt, copper, etc. The strong
affinity of iron for the non-metals explains its infrequent
occurrence in the native condition; and the dissimilarity
between the metal and its ores may explain why iron was
among the later useful metals to be discovered, if, as is gen-
erally believed, such is the case. It may be mentioned, how-
ever, that some writers think iron was known and used at a
much earlier period in the world's history than is generally
believed, but that the tendency of this metal to corrode has
destroyed all traces of its use in ancient times, while instru-
ments of brass and bronze remain.

Chemically pure iron is valuable only for experimental pur-
poses and as a curiosity, as it has no use in the arts except, per-
haps, in medicine. It may be obtained on a small scale in
several ways, among which may be mentioned the reduction
of pure ferric oxide by heating it in a current of hydrogen,
and the electrical decomposition of a solution of pure ferrous
sulphate or chloride.

While pure iron is devoid of value, when it contains small
quantities of other elements, it is the most useful and widely

§32

For notice of copyright, see pag^e immediately following the title page.

2 MANUFACTURE OF IRON §32

used of all the mejtals. In fact, it is almost impossible to
overestimate its importance in the arts.

It is not known who first discovered iron, nor is much
known of the early development of its manufacture ; and it is
not the object of the present subject to treat of the history of
the process, but to deal with conditions as we find them today.

The manufacture of iron from its ores depends on chemical
principles with which we are already familiar. As iron does
not occur native, it is necessary to reduce its compounds,
and this is done in such a manner that the resulting metal
shall contain the elements necessary to give it the properties
that have made it so valuable. The method almost univer-
sally employed at present is to charge in the ore, together
with the fuel — which at present is nearly always either coke,
coal, or charcoal — at the top of a tall furnace, and as the ore
always contains extraneous matter, a flux is also added in
the proper amount to form a fusible slag with these impuri-
ties. Hot air is blown into the furnace near the bottom, and
coming in contact with the highly heated fuel in excess,
forms carbon monoxide, which passes up through the descend-
ing charge of ore, fuel, and flux. At the temperature of the
furnace, both the carbon of the fuel and the carbon monoxide
thus formed act as reducing agents on the ore, removing the
oxygen and leaving metallic iron, which, at the intense heat
near the bottom of the furnace, melts and drops to the bot-
tom, taking up some carbon from the fuel, and silicon, sul-
phur, phosphorus, and manganese from the ore, fuel, and
flux. At the same time, the silica, alumina, lime, and mag-
nesia of the ore, fuel, and flux unite, forming a fluid slag,
which, being lighter than iron, floats on the molten metal in
the bottom. The iron and slag thus formed are drawn out
at proper intervals through openings provided for them in
the bottom of the furnace.

When the ascending gas reaches the top of the furnace, it
contains considerable carbon monoxide, which, as we already
know, is very combustible. It passes through an opening
near the top of the furnace and is led through the ** down-
comer " to a feedpipe. Part of it is conducted to the

§32 MANUFACTURE OF IRON 8

so-called stoves and burned in them to heat them up. The
stoves are then used to heat the blast of air blown in near
the bottom of the furnace by means of blowing engines.
The part of the gas not used in the stoves is burned under
the boilers that produce steam to run the blowing engines.

Having given a brief outline of the process, we will now
proceed to consider it more in detail.

IRON ORES

CliASSIFICATION OF IRON ORES

2. Deflnltlon of Ore. — In its generally accepted sense,
an ore is a naturally occurring substance containing a metal
in such quantity and condition that it may be profitably
worked for that metal. The metal may be either in the
native condition mixed with other substances or may be a
compound of the metal. As iron does not occur native, its
ores are compounds, and as only a few of these compounds
can be profitably worked for iron, the ores of iron are few in
number.

3. The oxides and carbonate of iron are about the only
compounds of this metal that are ores according to our defi-
nition. There are several of these, if we include combina-
tions of these oxides with other substances forming minerals
that are sometimes used as ores. It should be remembered
that no ore ever occurs pure, but always contains foreign .
matter in varying quantities. The impurities most fre-
quently found in iron ores are silica, alumina, lime, magne-
sia, manganese, phosphorus, and sulphur.

All ores contain some of these substances, and nearly all
contain all of them in varying proportions. Besides these
substances,^ which are nearly always present, a number of
others are found associated with certain ores. The follow-
ing are the principal ores of iron :

4 MANUFACTURE OF IRON §32

4, Magrnetlte. — Magnetite is an anhydrous oxide having
the formula Fefi ^^ and, consequently, if pure, would con-
tain 72.42 per cent, of iron, thus making it the richest of
the iron ores. It usually contains deleterious substances,
however, especially titanium, and frequently a high percent-
age of sulphur, and is not easily reduced, so it is not usually
considered as valuable as hematite, even though the per-
centage of iron in this ore is lower. Magnetite is not gen-
erally considered as a distinct oxide, but rather as a mixture
of ferrous and ferric oxide Fefi^^FeO, It is black, brittle,
and magnetic, and gives a black streak when drawn across
unglazed porcelain. It sometimes occurs in crystals and
sometimes in a granular condition, like sand, but generally in
the massive form.

6. Reel ITematite. — This is an anhydrous oxide having
the formula Fe^O^. It occurs in earthy and compact forms,
and a number of varieties are found, that is, crystalline, col-
umnar, fibrous, and amorphous. Special names have been
given to the various forms. Thus, the brilliant crystalline
variety is known as specular iron ore; the scaly, foliated
variety is known as micaceous iron ore, and the earthy varie-
ties are often known as red ocher.

This ore varies in color from a deep red to a steel gray,
but all varieties give a red streak when drawn across
unglazed porcelain.

Theoretically, it contains 70 per cent, of iron, and on
account of its abundance, its comparative freedom from
injurious constituents, and the character of the iron it
yields, it is the most important of the ores of iron.

6. BroMrn Hematite. — Brown hematite, or limonite,
is hydrated ferric oxide, and is generally represented by the
formula 2/r,(?„ 3 //,(?. Hence, it theoretically contains
59.89 per cent, of iron. It occurs in both compact and
earthy varieties. Pipe, or stalactitic, ore and bog ore are
brown hematite. Its color varies from brownish black to
yellowish brown, but it always leaves a yellowish -brown
streak on unglazed i)()rcelain.

§32 MANUFACTURE OF IRON 5

7» Ferrous Carbonate. — This ore has the formula
FeCO^ and thus, theoretically, contains 48.28 per cent, of
iron. It occurs in several varieties, known as spathic ore,
clay ironstone, and black band. Spathic ore, when quite
pure, has a pearly luster and varies in color from yellow to
brown. The crystallized variety is known as siderite.
When exposed to the action of air and water, the veins of
ore are decomposed to considerable depth and a layer of
brown hematite is formed. This ore frequently contains
considerable manganese, and in some places is used for the
production of spiegeleisen, which may be considered as iron
containing a high percentage (usually from 8 to 25 per cent.)
of manganese. Clay ironstone is a variety of ferrous car-
bonate that occurs in detached nodules or in layers of nod-
ules usually in the coal measures. It varies in color from
light yellow to brown, but the light-colored ore rapidly
becomes brown when exposed to air. Like spathic ore, it
usually contains considerable manganese. Black band is
a clay ironstone containing considerable carbonaceous mat-
ter, which gives it so dark a color that it frequently resem-
bles coal. The carbonates are not largely used as ores in
this country.

8. Pyrite. — According to our definition of ore, pyrite is
not an ore, in this country at least; but after extracting
the sulphur in the manufacture of sulphuric acid, the residue
of iron oxide, Jcnown as ** blue billy," is sometimes mixed in
small quantity with ores for the production of iron.

DISTRIBUTION OF IRON ORES IN THE

UNITED STATES

9» Magrnetlte. — This ore is found principally in a belt
running along the Eastern coast, from Lake Champlain to
South Carolina. There is considerable of it in New Jersey
and Eastern Pennsylvania, but the largest deposits are
found in Virginia and North Carolina. It is also found in
Missouri and in Northern Michigan, and is mined in East-
ern Canada.

6 MANUFACTURE OF IRON §32

The mineral franklinite is closely allied to this ore, and is
sometimes considered as a mixture of magnetite with the
oxides of manganese and zinc. It is generally considered as
a mixture of ferric and manganic oxides with ferrous, man-
ganous, and zinc oxides. In appearance it closely resembles
magnetite, but is less magnetic. In New Jersey, where it
occurs quite abundantly, it is treated for the extraction of
zinc, and the residue thus obtained is used for the manu-
facture of spiegeleisen.

10» Bed Hematite. — Until the discovery of the deposits
of this ore in the Lake Superior district, it was chiefly ob-
tained from a belt extending along the eastern coast of the
United States, just west of the magnetite deposits, and ending
in Alabama. Some of this ore is found in New York, but there
is not a great deal of it north of Danville, Pennsylvania. At
present, the greater part of the red hematite used in this coun-
try comes from the Lake Superior district. Ore of almost
any desired composition may be obtained in this district ; and
the enormous quantity of ore, its purity, the comparatively
small cost of mining, and the shipping facilities have made
this the great ore-producing section of the United States.

11. Brown Hematite, or liimonlte. — This ore is found
in a belt lying west of the red hematite in the eastern part
of the United States. Considerable of it was formerly
mined in Central Pennsylvania and there is much of it
in Alabama. It is also mined in the New England States
and in the Lake Superior district.

13* Carbonate. — This ore is important in Europe, espe-
cially in England, but there is not much of it in this country.
It usually occurs with bituminous coal or in the coal
measures. It is mined to a certain extent in Western Penn-
sylvania and Ohio.

Iron ore is found in several of the Western States, but as
these discoveries are comparatively recent, and as the mines
have not been developed to any great extent, it is impossible
to give much reliable information in regard to the ore in this
section at present..

§32 MANUFACTURE OF IRON 7

VAIiUATION OF IRON ORB

13. In deciding the value of an iron ore, several things
must be considered. Other things being equal, the value of
the ore will depend on the amount of iron it contains, and ore
is usually sold for a certain price per unit of iron. The
freedom of the ore from injurious constituents must also be
considered, and if considerable quantities of such substances
are present, the ore rapidly declines in value. The physical
properties of the ore and its proximity to market are also
important factors. As magnetite contains the highest per-
centage of iron of any of the ores, it would be the most
valuable of any, if the amount of iron alone were considered,
but as it usually contains considerable quantities of injuri-
ous substances— especially titanium and sulphur — and as it
is difficult to reduce, it is not as valuable as hematite. On
the other hand, the ore of the Mesabi Range, in the Lake
Superior district, is very rich and free from impurities, is
soft and easily reduced, and as it is a surface ore lying in a
horizontal layer, is mined very cheaply by means of steam
shovels. But it is not usually considered very valuable, for
it is very fine, and when charged into an ordinary furnace
running, as is usual, with other ores, much of it is blown
out with the escaping gases and it fouls the stoves and clogs
the boiler flues. The part that stays in the furnace tends
to hang to the walls for a time and then to slip, cooling the
furnace and producing poor iron.

PREPARATION OF ORES

14# In this country, most of the ores are used just as
they come from the mine, but in some cases a preliminary
treatment is an advantage and sometimes a necessity. The
preliminary treatment is very simple, however, and is usually
confined to three operations — viz., washing, crushing, and
roasting.

15. Washing:. — The rich ores, generally used at the
present time, require no treatment, but it is sometimes more

8 MANUFACTURE OF IRON §32

economical, on account of location or for some other cause,
to use poorer ores that may be improved by washing or
other treatment. For instance, the limonite ores that occur
in detached nodules mixed with clay are washed to remove
the clay, leaving the ore. This is usually accomplished
by first passing water over it in an ordinary trough, when
much of the finer material is carried off. The ore and the
remaining dirt are then carried to a revolving screen, known
as a trommel, and the remaining clay, etc. is washed out
by means of a spray of water. The arrangement of the
screen and the method of delivering the water are frequently
varied. Sometimes a pipe carries water in the middle of
the trommel, and sometimes a flat screen is used. In the
latter case, the screen is placed in an inclined position and
given a motion sidewise, the ore is run on to the upper end
of it, together with a stream of water, which washes the
clay through the screen, while the motion causes the ore to
pass on to the lower end of it.

16, Crushing:. — At the present time, most ores are
used just as they come from the mines, but some of the
hard, refractory ores that are mined in large lumps are
broken up before charging into the furnace. Probably the
form of apparatus most frequently employed for this pur-
pose is that of Blake's rock crusher, in which the ore is
crushed between a hard, moving jaw and a hard, fixed face.
Sometimes the ores are broken by hand, and stamps, rolls,
and centrifugal machines are used for this purpose. The
stamps used for this purpose are shoes of iron or steel having
an up-and-down motion, and acting on a steel plate. They
are usually worked in sets, each set consisting of several
stamps. The rolls are iron or steel cylinders, and are
usually worked in opposite directions. The centrifugal
machines are hollow cylinders containing large iron balls.
For uniformity in smelting, it is undoubtedly an advantage,
when refractory ores are used, to have the large lumps
broken up. The size to which the lumps should be reduced
will depend largely on the size and shape of the furnace

§32 MANUFACTURE OF IRON 9

and the character of the ore. Large lumps allow the gases
to pass through more freely, while small lumps or fine ore
pack more closely together and offer greater resistance to
the blast, thus increasing the pressure in the furnace, but a
larger surface is exposed to the action of the reducing gases.

17. Roastlngr. — Some ores are roasted to accomplish
one or more of several purposes. These are to desulphurize
the ore, either entirely or partially, to expel water, to expel
carbon dioxide, and to expel other volatile matter. Two
other objects that are not directly aimed at in roasting are
frequently accomplished. The ore is usually made more
porous, thus exposing a larger surface to the reducing gases,
and the lumps of some ores break up to a greater or less
extent. In the case of magnetite, roasting converts the fer-
rous oxide into ferric oxide, and thus lessens the liability of
the iron to pass into the slag.

Roasting or calcination is accomplished in open heaps, in
stalls, or in kilns. Where fuel is cheap and space is abun-
dant, ores are frequently roasted in open heaps. When this
method is adopted, a layer of coal a few inches thick is
spread on the ground, and a layer of ore is spread over it;
coal and ore are then added in alternate layers until the pile
is from 4 to 9 feet in height, the proportion of ore to fuel
increasing from the bottom towards the top. The coal at
the base of the pile is ignited and the combustion extends
throughout the pile. If at any time during the operation
any part of the surface indicates that the combustion is pro-
ceeding too rapidly at that point, it is damped down by the
addition of fine ore. The operation is allowed to proceed
until all the coal in the heap is burned.

Black band ore frequently contains enough carbonaceous
matter to accomplish the roasting without the addition of
fuel, except one layer of coal, which is placed on the
ground to start the combustion.

In some districts, calcination is accomplished in stalls,
which are rectangular spaces enclosed on three sides by walls
from 6 to 12 feet in height. These walls are perforated

MANUFACTURE OF IRON

§32

by two rows of air holes, each about 4 inches in diam-
eter. . The lower row is near the bottom and the second
row about 3 feet above the first. The floor of the stall
usually slopes slightly towards the open side. The operation
of roasting is conducted the same as in open heaps, but less
fuel is used, the draft is more under control, and a more
perfect calcination is accomplished.

The calcination of ores in kilns is more economical, as
regards fuel and labor, than either of the foregoing proc-
esses, and, in addition to economy, the process is more
under control and a more uniform product results. There
are a number of forms of kilns. Among the best known are
the Gjers, Grittinger, and Davis-Colby.

In Fig. 1 is shown a vertical section of the Gjers calcining
kiln. This is a circular kiln built of iron plates and lined

with about 14 inches
of firebrick. A com-
mon size for kilns of
this kind is 14 feet in
diameter at the bot-
tom, 20 feet at the
widest part, and 18
feet at the top. The
height of such a kiln
would be about 30 feet
and its capacity 6,000
cubic feet, but they
are also constructed of
more than twice this
capacity. The kiln
rests on a cast-iron
ring supported on cast-
iron columns about 30
inches in height, leav-
ing a clear space for the

vV,, •.•C\.,.

s\S\\Vs\\N^N\\\V<^;NNv N\>;ii^-i>»»""*'''^^

Fio. 1 removal of the roasted

ore. A cast-iron cone with its apex upwards rests on the
ground in the center of the kiln. This directs the descending

§33 MANUFACTURE OF IRON 11

13 MANUFACTURE OF IRON g 32

ore outwards and renders its removal more easy. A series
of openings d supplied with doors extends around the kiln
near the bottom, for the admission of air; and the supply
may be regulated by opening or closing doors as the case
may demand. As the roasted ore is removed from the
bottom, fresh ore and fuel are added at the top; hence, the
process is continuous.

The Grittinger ore roaster is shown in Fig. 2. Like the
Gjers kiln, it is built of iron plates lined with firebrick. The
kiln rests on a masonry base. A star-shaped cone rests on
the masonry in the bottom of the kiln, and this directs the

ore outwards into the chutes. Fig. 3 shows the star-
shaped cone and the chutes as they appear when looked at
from above. A large flue passing up in the cone supplies
air to the center of the kiln, and the openings near the bottom
supply air to the outer portion of the ore. The roaster
shown in Fig. % is supplied with a cast-iron hood and

§32

MANUFACTURE OF IRON

chimney. These are not essential parts of the roaster, but
are advantageous when ores containing much sulphur are
roasted, as they carry off the fumes of sulphur dioxide. The
raw ore and fuel are charged at the top, as in the Gjers kiln,
and the roasted ore is discharged from the chutes.

The Davis-Colby ore roaster, shown in Fig. 4, uses gas as
the fuel. It consists of two concentric shafts of brickwork
having between them a space ^, about 18 inches at the top
and 24 inches at the
bottom, to contain
the ore under treat-
ment. The outer
shaft contains the
gas flues x^ the fire
arches b^ the air flues
and poking holes z^
and the chutes r, for
removing the roasted
ore. The inner shaft/
contains the open-
ings d through which
the products of com-
bustion enter the
shaft. These waste
gases are carried
down and escape
through the under-
ground flue f\ which
connects with the
draft stack. The ore
is dumped upon the
cone /, which covers
the inner shaft, and
this distributes it.
This roaster is sometimes modified by building the draft
stack on the shaft /and leading the products of combustion
up through this, but when this is done the difficulty of
charging the ore is increased.

Fio. 4

14 MANUFACTURE OF IRON §32

18, Rules for Desulpliurlzlnfir Ores. — The following
conclusions are drawn, from the experiments conducted by
Mr. Valentine at Lebanon, Pennsylvania.

1. Heat alone, without access of air, can remove at best
only one-half of the sulphur present.

2. Atmospheric oxygen is absolutely necessary for a
proper desulphurization.

3. Even at a low heat, ore is properly desulphurized if
air can gain access freely to the FeS^ in it.

4. Sulphate of iron can be decomposed by heat equally
well with or without air.

5. In order that the residuum of sulphur in roasted ores
may consist, so far as possible, of sulphates, the roasting
must be done under free access of air.

6. Fusion or sintering of ore is likely to prevent any
further desulphurization.

7. Sintering does not allow much of the remaining sul-
phur to be in the form of sulphate.

8. Fusion, hence, should never occur in roasting, except
after continued heating in air at a lower temperature.

9. Ores cannot be properly desulphurized in the upper
part of the blast furnace.

10. An efficient roaster must allow easy control of heat,
abundant access of air to the hot ore, and rapid removal of
the products of combustion.

FUEIi

19. Quite a variety of fuels may be used in the blast fur-
nace, provided the furnace is modified to suit the particular
case. With proper modifications, it has been found that
raw non-caking coal, turf, and wood are available, but in
this country at the present time the only fuels used to any
considerable extent are coke, charcoal, and anthracite coal.
Coke is, on the whole, the most satisfactory fuel for the blast
furnace, and is much more largely used than either of the

§32 MANUFACTURE OF IRON 15

others. Charcoal is used to a certain extent on account of
its freedom from impurities and because it is generally
believed that charcoal iron is better for some purposes than
the iron made in a coke furnace. Anthracite is used princi-
pally in Eastern Pennsylvania, as the proximity of the mines
makes it the cheapest fuel available. In some cases a mix-
ture of coke and anthracite is used.

30. Coke. — Coke is the combustible residue left when
the volatile constituents of bituminous coal are expelled by
heat. Much of the sulphur of the coal is expelled with
volatile hydrocarbons, so that the coke usually contains con-
siderably less sulphur than the coal from which it is made.
Coke made from different coals and by different methods
varies both in composition and properties. In determining
the value of coke as a blast-furnace fuel, both the chemical
composition and the physical structure must be taken into
account. A good coke for furnace use should contain a low
percentage of ash, sulphur, and phosphorus and a high per-
centage of fixed carbon.

It should be strong and hard, for the softer coke softens
still more when heated in the upper part of the furnace and
will not bear the weight on it well. In addition to this, it
burns more readily than the hard coke, and much of its
power is gone before it reaches the bottom of the furnace,
where it should burn in order to work economically. At
all events, the coke should be uniform in composition and
properties, as it is impossible to produce good results with-
out a uniform fuel.

21. Charcoal. — Charcoal is the carbonaceous residue
that remains when wood is partially burned with a limited
supply of air, or heated out of contact with air. It is generally
obtained by making a pile of wood closely packed together,
covering it with earth to allow but little air to come in contact
with it, and igniting the wood. Part of the wood is thus
consumed in charring the remainder. Considerable charcoal
is also obtained as a by-product in the manufacture of wood
alcohol and other manufacturing processes.

16 MANUFACTURE OF IRON §32

Charcoal is more expensive than coke, and as it is not as
strong, it will not bear up the burden as well; hence, it is
only used in comparatively small furnaces. On account of its
freedom from impurities, it is thought to produce a superior
grade of iron, and is consequently used to a certain extent.

22, Anthracite. — Anthracite is a strong hard coal and
bears up the burden well while at a comparatively low temper-
ature, but as it is very dense it burns slowly, and when it
comes to the hotter portion of the furnace it decrepitates, fall-
ing into small pieces that cause the charge to descend slowly
and hinder the ascent of the gases. Hence, a higher blast pres-
sure is generally used in furnaces using anthracite than in
those using coke. The anthracite furnace should also have a
larger diameter in proportion to its height than a coke furnace.

FliUXES

23. As nearly all iron ores contain an excess of silica, a
basic substance that will unite with the silica, forming a
fusible slag, is required as a flux. For this purpose, lime-
stone is almost universally employed, though dolomite is
used to some extent. The value of a limestone as a Bux
depends on its freedom from impurities, especially silicon
and sulphur. The presence of silica in the stone rapidly
reduces its efficiency as a flux, and as the lime unites with
the sulphur of the stock, thus removing it from the furnace
in the slag, it is important that the stone should be free from
sulphur to start with. A small amount of magnesia in the
stone appears to be an advantage, but a high percentage is
detrimental, except in the production of spiegeleisen, ferro-
manganese, etc., in which cases a difficultly fusible slag is
desired. For the production of Bessemer iron, it is also
important that the stone should be free from phosphorus,
for practically all the phosphorus in the stone, as well as that
in the fuel and ore, go into the iron. A small amount of
magnesia in the flux appears to make the slag more fusible,
but a larger quantity decreases its fusibility; and, conse-
quently, a stone containing but little of it is preferred.

§ 32 MANUFACTURE OF IRON 17

The carbon dioxide is driven off by the heat before the
stone begins to act as a flux, and an attempt was made to
economize fuel by burning the stone to lime, using a cheap
fuel for this purpose, before charging it into the furnace.
This was not successful, however, for the lime appears to
absorb carbon dioxide from the escaping gases in the top of
the furnace, and is changed back to carbonate. The carbon
dioxide must then be driven off again in the hotter part of
the furnace before fluxing begins.

A few ores contain an excess of basic material, and in such
cases an acid material must be added to form a slag. This
is frequently done by mixing with a silicious ore in the
proper proportion.

)84, The Efflelency of Ijimestone. — The usual method
of calculating the efficiency of a limestone is as follows:
Multiply the percentage of lime by .54; multiply the per-
centage of alumina by .87; multiply the percentage of mag-
nesia by .75. Add the results, subtract the percentage of
silica, and the result will be the amount of silica that the
stone will flux. For example, take a stone having the com-
position

SiO^ = 5.00j^

CaO = 50.64^
MgO = 1.27^

Then, .80 X .87 = .70

50.64 X .54 = 27.35
1.27 X .75 = .95

29.00

This gives the amount of silica that will be fluxed by the
basic material in the stone. Then, subtracting the silica in the
stone, we have 29 — 5 = 24, the efficiency of the stone in terms
of silica, or, in other words, 100 pounds of the stone will flux
24 pounds of silica contained in the ore and fuel. This method
of calculation of the efficiency of a stone is useful in many
cases, but it is based on theoretical calculations, and in
practice yields a slag that is too acid for ordinary purposes.

18 MANUFACTURE OF IRON §32

BTX>WTNG ENGLNES

25. The details of the blowing engines belong to the
province of the mechanical engineer rather than to that of
the chemist or nietalhirgist, but as these engines are of vital
importance in running a furnace, a brief description will be

given. The blowing engines at most of the furnaces at the
present time are of the vertical type, similar to that shown
in Fig. 6. There are a number of these engines that have
proved themselves very good, each having some advantage
over the others.

§32 MANUFACTURE OF IRON 19

In these engines the blowing cylinder, or ** blowing tub,"
as it is usually called, is at the top. It is fitted with valves,
so that when the piston passes up the air is forced out at the
top to the blast main, and at the same time air is drawn in
at the bottom of the cylinder. When the piston passes
down, air is forced out of the bottom of the cylinder to the
blast main, while air is drawn in at the top to fill the
cylinder. A good arrangement is to have at least three
blowing engines, any two of which will easily produce all
the blast ever required. Then, by always having one engine
idle and alternating, all necessary repairs may be made with-
out interfering with the blast, which is so important for the
successful working of the furnace.

STOVES

36. Formerly a cold blast was used in furnaces, but con-
siderable extra fuel was required to heat the large amount
of air blown into the furnace, and at present a hot blast is
used almost exclusively. The blast is heated in so-called
stoves by means of the waste gases of the furnace. These
stoves are of two kinds, viz., iron pipe stoves 3,nA regenerativi
stoves.

PIPB 8TOVB8

37. When the hot blast was first introduced, cast-iron
pipe stoves were employed to heat it. One of the earliest
forms of stoves is shown in Fig. 6. It consists of an oblong
chamber of firebrick, along each long side of which circular
mains a and b pass near the bottom of the chamber. These
mains are fitted with sockets that receive the ends of
inverted U-shaped cast-iron pipes, which form an arch and
connect the two mains. Each stove usually contains from
eight to twelve of these U-shaped pipes. The gas from the
furnace is burned in the chamber between the mains, and
the flame passes up between and around the pipes, thus

20 MANUFACTURE OF IRON § 32

heating them. There are generally partitions in the mains
between the sockets, so that the cold air forced in. a passes
through the first pif>e to *, then back through the second
pipe to a, and so on
until it has passed
through all the pipes
and has become
heated, when it
leaves the stove by
the hot-air main.

This stove has been
modified in many
ways until at present
there are a number
of forms of pipe
stoves in use, but all
depend on the same
principle and may be
considered as modi-
fications of the stove
just described.
These stoves are
much cheaper than
the regenerative
stoves, and produce
a comparatively even
temperature, which
is favorable for the
production of a good
quality of foundry
iron, and conse-
quently are well liked by many furnacemen at small furnaces
making foundry iron. The principal objeclions to them are
that they will not heat the blast to a temperature exceeding
1,000° F. without rapidly burning out the pipes, and if an even
temperature is not continually maintained the pipes crack
from the expansion and contraction caused by changes in
temperature, and the expense of keeping the stoves in repair

§32 MANUFACTURE OF IRON 21

is relatively large. In addition to these objections, the back
pressure in pipe stoves is always considerable and there is
always more or less leakage, thus throwing extra work on
the blowing engines.

REGENERATIVE STOVES

38. On account of the objections just enumerated and
the relatively high temperature of blast now employed, the
pipe stoves have been quite generally superseded by stoves
built on the regenerative principle of Sir W. Siemens. The
adoption of these stoves has been attended with considerable
saving in fuel and an increased output, and the cost of
repairs has been greatly reduced. There are four types of
regenerative stoves, viz., the Cowper, the Whitwell, the
Massick and Crooks, and the Kennedy stoves. These are
all good and are all largely used.

29. The CoMrper Stove. — This stove is round in form
and consists of a wrought-iron casing lined with firebrick.
It is covered with a dome-shaped roof, also lined with fire-
brick. Most of the interior of the stove is filled in with a
checkerwork of firebrick, but at one side of the stove a cir-
cular flame flue, or combustion chamber /, is left clear.
Fig. 7 is a vertical section showing the flame flue f and
Fig. 8 is a cross-section showing a general plan of the stove.
The furnace gas enters the stove through the valve g^
mingles with the air that enters through the valve ^, and
burns in the combustion chamber /. The divisions in the
combustion chamber, shown in Fig. 8, are to secure a more
thorough mixture of gas and air. The flame and heated
products of combustion pass up under the dome and then
down through the numerous passages in the checkerwork of
firebrick, which is supported by the pillars /. In their
downward passage, the gases give up much of their heat to
the checkerwork of brick and finally pass out of the stove at
5, into the draft stack. After burning the gas in the stove
for a time, the brickwork becomes highly heated, especially

MANUFACTURE OF IRON

§33 MANUFACTURE OF IRON ^ 83

near the top of the stove. The valves a, g, and s are now
closed, and the cold-air valve c near the bottom of the
checkerwork and the hot-air valve near the bottom of the
combustion chamber are opened. The cold air from

the engines enters near the bottom and, passing up through
the checkerwork of hot brick, becomes heated. It then
passes down through the combustion chamber and out at
the valve h to the hot-blast main leading to the furnace.
At the sides of the stove are the cleaning doors d.

30. The Whltwell Btove This, like the Cowper stove,

is a tall, round stove consisting of a casing of wrought iron
lined with firebrick. It differs from the Cowper stove prin-
cipally in the arrangement of the inner brickwork and in
the method of admitting the air to burn the gas. Fig. 9
shows a vertical section of this stove. The furnace gas
passes into the stove through the valve ^, where it meets a
limited supply of air, introduced through a, and partly burns

§32 MANUFACTURE OF IRON 25

as it passes up through the flame flue, or combustion cham-
ber y. The unconsumed gas and the products of combus-
tion pass down through the narrow chambers, as indicated
by the arrows. At the bottom of the stove, the gas meets a
fresh supply of air, introduced through a\ and the combus-
tion is completed \r\f\ The hot products of combustion pass
down through the narrow chambers, as indicated by the
arrows, and escape through the valve s to the draft stack.
When the brickwork is thoroughly heated, the valves ^, g^
and s are closed, and the cold-blast valve c and hot-blast
valve h are opened. The blast passes through the stove in
the reverse of the course taken by the furnace gases and
passes out by the valve It to the hot-blast main.

As we have seen in the Cowper stove, the gas and the air
pass out of the stove after passing up and down through the
stove once, while in the Whitwell the gases and the air each
pass up and down twice before leaving the stove ; hence, the
Cowper stove is called a two-pass stove and the Whitwell
a four-pass stove.

31. The Massick and Crook's Stove. — This stove has
the combustion chamber, or flame flue, in the center. The
hot products of combustion pass down through chambers
just outside of the combustion chamber, and pass up through
chambers next to the walls of the stove, escaping through a
chimney built on top of the stove. It is therefore a three-
pass stove.

32. Tlie Kennedy Stove. — The Kennedy stove is built
much like the Whitwell, but differs from it in the arrange-
ment for burning the gas and the passage of air. In all
the stoves mentioned thus far, the combustion chamber is
heated very highly, while the portions of the stove traversed
by the products of combustion just before their escape to
the chimney are relatively cool. Mr. Kennedy sought to
equalize the temperature by admitting gas and air at the
bottom of the four chambers, thus burning the gas as it
passes up through these chambers and allowing the products

26 MANUFACTURE OF IRON §3^

of combustion to escape directly by a chimney on the top
of the stove, thus having a direct natural draft. When the
stove is heated, the valves are changed, and the blast passes
through all four chambers, thus making the stove one-pass
for gas and four-pass for the blast.

33. General Remarks on Stoves. — All these stoves
have been modified to a greater or less extent, to suit the
conditions at different furnaces. In many places, instead
of admitting gas and air by separate valves, the gas is
introduced into the stove through a jet pipe carried on a
horizontal slide covering the opening in the gas main. From
this it is blown into a circular opening in the stove, which is
larger than the jet pipe. The gas thus forced in under
pressure draws in the air necessary for its combustion. At
each furnace, there should be three or four stoves of suffi-
cient size to heat the blast. If there are four stoves, the
blast will be passing through two of them while two are
being heated, and by alternating at frequent intervals an
even temperature may be maintained.

THE FURNACE

34. There have been great changes in the size and
form of furnaces in the last half century, and as changes
are continually being made in the style of building, it is
impossible to give the most favorable dimensions, for this
is a matter that has not been determined. Indeed, if fur-
naces of several types were run side by side for some time,
this would not establish the most favorable form for use
under all conditions, for the form of furnace that will
produce the best results with one kind of stock will not
work well with another kind. For instance, a tall furnace,
relatively small in diameter, is now generally considered the
best for coke practice ; but if charcoal were used in such a
furnace, it would be crushed by the great weight of stock.
Anthracite, though strong, decrepitates when strongly
heated in the furnace, and fine particles mixing with the

28 MANUFACTURE OF IRON §33

slag as it is forming produce a compact mass through
which it is difficult for the blast to penetrate for any con-
siderable distance; hence, a furnace in which anthracite is
to be used as the fuel should be rather low and relatively
large in diameter. Fig. 10 shows a form and size of fur-
nace that appears to be popular at the present time for coke
practice, as several furnaces of almost exactly the same
dimensions are being built in this country, or have recently
been completed. The lower part of the furnace, known as
the hearth, or crucible, is 13 feet in diameter and 9 feet
6 inches deep. The walls are built of the most refractory
firebrick, to withstand the intense heat of the molten iron
and slag that collect here.

From this point the diameter steadily increases up to the
mantel, 15 feet above the top of the hearth, where the
diameter is 21 feet. The part of the furnace from the hearth
to the mantel is known as the bosh or boshes. From here
up the diameter decreases regularly until it is 14 feet at
the stock line. The throat is 70 feet 6 inches above the
mantel. The stack, as the part of the furnace above the
mantel is called, rests on strong columns of iron that are
set firmly on the foundation. The stack is encased in
wrought-iron plates, which are firmly riveted together.
Though the tendency at present is to build large furnaces,
most of the furnaces in this country are smaller than the
one shown in the illustration, but those recently constructed
resemble the furnace shown more or less closely in general
form.

36, Protection to Furnace liiningrs. — The heat near
the bottom of the furnace is intense and the stock descend-
ing, combined with the heat, tends to wear on the brick
lining. In addition to this, the slag tends to attack the
lining, thus wearing it away still more rapidly. To protect
the lining so far as possible, hollow plates/. Fig. 10, are set
in the brickwork and a current of cold water is kept flowing
through them. These plates are made of cast iron, wrought
iron, bronze, and copper, but the copper plates appear to

§32 MANUFACTURE OF IRON 29

wear better than the others... The cold water coming close
to the inside of the lining cools the bricks and causes a thin
layer of slag to solidify on them, which protects them from
the further action of the slag. A number of rows of these
are set in the lining, completely surrounding the furnace.
As a rule these coolers have been set in up as far as the
mantel, and in furnaces thus protected it has been noticed
that after running some time the hearth and boshes were in
very good condition, but that the brickwork just above the
upper row of plates was worn back for some distance, form-
ing an offset in the lining. To remedy this, two or three

•

rows of plates have been set in above the mantel in some of
the new furnaces. The furnace shown in Fig. 9 has two
rows of these plates. The plates above the mantel are gen-
erally set back about 1 foot from the inside surface of the
lining, to allow the furnace to assume the most favorable
working lines. The lining near the stock line — the point
near the throat to which the stock extends when the
furnace is working — is usually worn quite rapidly by the
coarse stock falling or rolling against it. To prevent
this, brick-shaped cast-iron plates were set in the lining of
some furnaces at this point. They protected the lining at
this point, but were very heavy. More recently a casting
having the shape shown in Fig. 11 has been used in some
furnaces, and answers the purpose
remarkably well. The advantage
of this form over a brick-shaped

casting is that the lining receives ^ . ,^'.»S

equal protection, while the weight is

greatly decreased. The furnace is

usually surrounded by a wall with suitable openings, and a

roof is built around the furnace above the mantel, thus

forming a kind of rude house to protect the men working

around it from the weather and from pieces of stock falling

from the top of the furnace.

36. Tuyeres. — The hot blast as it leaves the stoves
passes through a large pipe lined with firebrick to the bustle

MANUFACTURE OF IRON

5 33

pipe b. Fig. 10, which is also a large pipe lined with firebrick
running around the furnace and generally supported by
brackets on the columns that support the stack. From this
the blast is carried to the tuyeres t. Fig. 10, by means of
pipes. The tuyeres are set in the wall of the hearth and
extend through, as shown in the figure. They are thus sub-
jected to extremely destructive influences, for besides the
blast heated from 800° to 1,400° F. passing through them, the
inner ends come in contact with molten metal and slag, and
the heat of this part of the furnace is intense. The tuyeres
are made of cast iron, wrought iron, bronze, or copper, and
are always cooled by water. The method of cooling varies
somewhat; one of the older forms was a hollow, truncated
cone, through which a constant current of water was kept
flowing between the
opening for the blast
, and the outside by
means of supply and
exit pipes. A form
that is more common
at present is shown
in Fig. 12. A spiral
pipe runs through the
part of this tuyere
^_, that was left hollow in

the older forms, and
a current of water is
maintained through
the spiral. Copper
tuyeres last better
than those of iron,
for besides standing
the high temperature
better, they are not
attacked by the par-
tially fused masses of
iron that frequently
adhere to iron tuyeres

§32 MANUFACTURE OF IRON 31

when the furnace is working badly. The number and size
of tuyeres vary with the size of the furnace.

In the furnace shown in Fig. 10 there are ten, which is
a common number in the larger furnaces. As many as six-
teen are sometimes used. The tuyeres are usually placed in
a horizontal plane. If they dip downwards, it is said there is
danger of the blast playing on the surface of the molten
metal in the crucible, thus decarburizing it and producing
white iron. It is said that sometimes in making gray iron
there is an advantage in directing the nozzles slightly
upwards, but this can scarcely be said to be proved.

A row of blank tuyere openings are usually built into the
furnace wall above the tuyeres ordinarily used. In case the
hearth partly fills up when the furnace is working badly,
these may be broken through and tuyeres inserted. They
are usually called monkey tuyeres.

37. Iron and Cinder Notc^hes. — The iron notch, or tap
hole^, Fig. 10, is generally an oblong opening, near the bot-
tom of the hearth, lined with cast iron. It is properly cooled
by water that is generally led through a spiral pipe. The
opening through the iron is closed with clay, or if the clay
at hand is too silicious, with a mixture of clay and coke dust.
At proper intervals a hole is drilled in this clay for the iron
accumulated in the hearth to pass out. The hole is drilled
downwards, entering the furnace near the bottom of the
hearth and practically all the iron is forced out by the blast.
When the hearth is empty, the blast is turned off and the
hole is closed with clay. This is hardened almost immedi-
ately by the heat, and the blast can be turned on as soon as
the hole is closed. The number of casts made in a day will
depend on the size of the hearth, the rate of driving, etc.
When the furnace is working regularly, there is usually a set
time for each cast and five or six casts are usually made in
24 hours.

Considering the part of the furnace from which the iron is
tapped as the front, the cinder notch c, Fig. 10, is usually
situated at the side, that is, one-fourth of the distance around

32 MANUFACTURE OF IRON §38

the furnace from the iron notch. It is situated on a level
between the tap hole and tuyeres. In modern furnaces,
the cinder notch resembles a tuyere and is cooled by water
in the same manner. When the slag has run out, the open-
ing is closed by a piece of metal — usually bronze— on the end
of an iron bar. This chills the slag, which solidifies, effectu-
ally closing the hole in a few moments, and the bar may be
withdrawn. When it is desired to remove the slag, that
chilled in the inner part of the hole is easily broken through
with a bar. The slag is usually conducted through a trough
to a slag car or ladle, in which it is hauled to a slag dump, or
cinder dump, as it is usually called. The slag is withdrawn
more frequently than the iron, the frequency depending
somewhat on the working of the furnace. As a rule, from
three to five flushes are made between each cast and the
succeeding one.

38, Bell and Hopper. — Before the blast-furnace gases
were utilized as fuel, the throat of the furnace was left
open, and a chimney was usually built to carry off the
gases. At present, furnaces are closed and the gas is col-
lected. The device by which the throat is closed is known
as the bell and hopper, and is illustrated in Fig. 13. This
differs in details in different places, but consists essentially
of an inverted truncated cone, known as the hopper, set in
the throat of the furnace. Beneath this is suspended a
cast-iron cone known as the bell. The bell may be raised
or lowered ; but when raised, the joint between it jand the
hopper must be tight. When lowered, an opening is left
between it and the hopper, through which the stock passes
into the furnace.

39, The Downcomer. — Just below the hopper an open-
ing is left in the wall of the furnace, through which the gas
passes to the downcomer. This is a pipe leading down
almost to the ground. At the lower end it is enlarged,
forming what is known as the dust catcher. This, as its
name indicates, is designed to collect the fine stock, etc.

§ 33 MANUFACTURE OF IRON 33

carried out of the top of the furnace by the gas. It is open
at the bottom, and is fitted with a small bell similar in form
to the bell at the top of the furnace.

When the blast is stopped after each cast, while the tap
hole is being closed, the bell is lowered and the dirt that has
accumulated is allowed to drop out. An opening in the top
of the dust catcher connects with the gas main, which runs
past the stoves and to the boilers. Enough of the gas is
burned in the stoves to heat them, and the rest is burned
under the boilers that produce the steam to run the blowing
engines, pumps, etc. If there is gas enough, no solid fuel
need be used under the boilers while the furnace is running
properly, but, as a rule, it is necessary to burn coal under
the boilers in connection with the gas.

40. Explosion I>oor8. — The stock in a certain part of
the furnace sometimes stops in its descent, or hangs, as it is
called, until that beneath It has passed down some distance,
and then slips down. This frequently causes a more or les;s

MANUFACTURE OF IRON

§32

violent explosion in the furnace, which in extreme cases
would wreck the top of the furnace if no means were pro-
vided to relieve the sudden pressure. To provide for such

PIO. t4

cases, openings are made at the top of the furnace, and
these are closed with doors held in place firmly enough to
resist the pressure of the blast, but will be forced open and

§32 MANUFACTURE OF IRON 86

relieve the pressure in case of an explosion. Such a door is
frequently placed at the top of the downcomer, and one on
the opposite side of the furnace, though this arrangement is
by no means universal.

41. Cliargrinsr. — At most of the furnaces erected some
time ago, and at many recently built, the stock is raised to
the top of the furnace in hand barrows by means of vertical
hoists. The barrows are then dumped in the hopper by
hand, and an even distribution of ore, coke, and limestone
is thus easily obtained. Later, at some furnaces, sloping
hoists were built, and small cars were run to the top of
the furnace on rails and dumped by mechanical means. The
earlier forms of this hoist were not very satisfactory, as the
stock was not evenly distributed by them, and the furnace
was thus caused to work unevenly. Quite recently, how-
ever, a number of furnaces have been equipped with
improved mechanical charging devices that have given gen-
eral satisfaction. One of these devices is shown in Fig. 14.
The stock is dumped from the hoisting car c into the small
hopper A, from which it passes through the chute c' to the
regular hopper beneath. As the car goes down and comes
up again, this small hopper makes part of a revolution, so
that each succeeding car of stock is dumped in a different
part of the hopper. It is so arranged that it may be set to
dump any desired number of times in making the circuit of
the top of the furnace, and an even distribution of stock is
thus made possible. At proper intervals the bell is lowered
to allow the stock in the hopper to fall into the furnace,
where the surface of the stock takes the form of the line s.

BliOWING IN THE FURNACE

43. Formerly much time was spent in blowing in a
furnace and getting it to running regularly, but at pres-
ent this is accomplished much more rapidly. • The method
adopted varies at different furnaces. Some wood is nearly
always used in blowing in, but the amount varies, and in

36 MANUFACTURE OF IRON §38

most places less wood is now used than was formerly cus-
tomary, as its use is in some ways objectionable. It con-
tains but little matter that can be fluxed oflf, and in some
cases some of it has charred and formed lumps on the wall
that remained there when the furnace was blown out for
repairs. The same objection holds with respect to charcoal.
In some cases wood is placed in the hearth, while in others
a scaffold is built up about to the tuyeres and the wood is
placed on this. In some cases, one or two rows of cord
wood are stood up around the walls of the furnace above the
wood to protect the lining. At present, slag is frequently
added with several of the first charges put in the furnace,
the amount gradually diminishing with the succeeding
charges. When the furnace is lighted, this slag melts and
runs down into the hearth before the ore farther up in the
furnace is reduced, thus heating the hearth and preparing
it for the iron. It is a good plan in blowing in to so pro-
portion the limestone that the first slag will be slightly acid,
for a basic slag attacks the lining much more rapidly than
an acid slag, until a coating of slag and graphitic material
has formed on the lining. A very acid slag should be
avoided, however, as this wears the lining quite rapidly.

A method of blowing in that has proved very satisfactory
is to place coke in the crucible up to within a couple of feet
of the first row of coolers. On this is piled wood — generally
cord wood and dry pine broken up rather fine. The wood
in front of the tuyeres is saturated with oil. One or two
tiers of cord wood are now frequently built up around the
walls to protect the lining. About 20 tons of coke are next
added and then sufficient limq^tone to flux the ash of the
coke, and an equal weight of slag. Generally, about 1 ton
each of limestone and slag will be the proper amount. If
the furnace is small, a little ore may be added with the next
fuel ; but if large, this charge should be repeated. Above
this, several charges are added, each containing about one-
fourth as much ore, by weight, as coke and sufficient lime-
stone to flux the silica of the ore and coke, together with a
weight of slag equal to that of the limestone. The weight of

§32 MANUFACTURE OF IRON 37

slag added now decreases with each charge, and after a
few more charges is discontinued. The weight of ore and
limestone, on the other hand, is steadily increased, until the
proportion at the stock line is about one of coke to one and
one-fourth of ore, and sufficient limestone to flux the silica
of the ore and fuel. The furnace is now lighted at each of
the tuyeres. This may be done in several ways. A good
method is to run a red-hot bar through each tuyere, thus
igniting the oil. Waste saturated with oil is sometimes
placed in front of each tuyere for this purpose. When
ignited, a gentle blast heated to about 600** F. is turned on ;
this is gradually increased until in a short time about
one-fourth the blast generally used is being employed. The
blast is heated by passing it through stoves that have pre-
viously been heated by burning coal, wood, or coke in their
combustion chambers and using a gentle draft.

As soon as the wood bums out and the coke settles down
in front of the tuyeres, carbon monoxide is formed. The
bell should be left open until this burns steadily at the top
of the furnace. The bell is then raised and the gas is
usually led to the boilers first. After burning here for a
short time, it is used in the stoves in the usual manner.
The furnace is kept full by adding fresh stock as fast as
that charged in before lighting settles. Soon after the coke
commences to burn, the slag charged in will begin to melt
and trickle down to the bottom of the hearth, and slag will
be formed by the union of the limestone and ash of the coke.
The hot slag collecting in the hearth heats it up, and the
temperature of the hearth is further raised by the coke
added below the wood burning here. When considerable
slag has collected in the hearth, it is withdrawn through the
iron notch. This is repeated several times until the iron
begins to collect, and then the cinder notch is used when-
ever it is necessary to withdraw the slag. Any iron that
may have collected and passed out with the slag is sep-
arated from it and returned to the furnace. The slag that
collects in the hearth at first and is withdrawn through the
iron notch heats up the hearth and clears it out, thus

38 MANUFACTURE OF IRON §32

preparing it for the reception of the iron, which soon begins
to collect. After the blast has been turned on for a few
hours it is increased, from time to time, until in a few more
hours, if everything goes right, the blowing engines will be
running at the ordinary rate. The experienced furnaceman
can readily tell from the condition of the furnace how rap-
idly it is safe to increase the blast. The burden is also
increased as rapidly as conditions will warrant, until the ore,
fuel, and limestone are being added in the usual propor-
tions. With careful handling, the furnace should be run-
ning as usual in a few days after blowing in.

BliOWING OUT

43. After the furnace is blown in, it is run continu-
ously— unless it is necessary to stop a short time for repairs
— until it is necessary to suspend operations, in order to
reline the furnace, or for some other reason. The work
may be stopped for a few days by first adding enough extra
fuel to make up for the loss of heat during the stop and then
closing the furnace, so that no air can get in ; but if a pro-
longed stop is necessary, the contents of the furnace must
be removed. This is known as blowing out. The blowing
out of a furnace, like the blowing in, is accomplished in
several different ways. A method that is being employed
quite largely at present, because it protects the top of the
furnace, is as follows:

When the furnace is running normally stop the addition
of ore and continue to charge fuel with just enough lime-
stone to flux the ash for 8 or 10 hours; then add fuel alone,
keeping the furnace filled to within 10 or 15 feet of the top.
When the ore is all reduced and the coke and limestone
begin to enter the hearth, an excess of gas will be produced.
Some of this is allowed to escape and burn at the top of the
furnace by opening the bleeder. Later, it will probably be
necessary to open the explosion doors also. Continue to run
as usual for from 3 to 5 hours for any difficultly fusible
material in the hearth to melt and then gradually reduce

§3a MANUFACTURE OP IRON 3d

the blast. Finally drill a hole through the iron notch as low
in the hearth as possible and blow out the last of the molten
material. Now shut oflf the blast and close the tuyeres with
clay. After standing for 24 or 36 hours, remove two or
three of the tuyeres and rake the coke out through these
openings. A stream of water is directed on the coke as it
falls to the ground in front of the tuyere openings, to cool
it. It is then taken to the stock pile to be used again.
When the coke is raked down level with the tuyeres, the
remainder is cooled with water, and when sufficiently cool,
workmen are sent in to finish cleaning out the hearth. By
keeping the furnace pretty well filled with coke in this way,
while the last of the ore is being reduced, the top is pro-
tected from the intense heat of the hearth.

CASTING

44, When considerable iron has collected in the hearth,
a hole is drilled nearly through the clay that closes the iron
notch and a bar is driven through the remaining portion.
The weight of the iron causes much of it to run out, and the
pressure of the blast causes it to run faster and forces out
the last portion remaining in the hearth, together with the
slag that has accumulated, forming a layer on the iron.
The iron runs along a trough made in the sand, known as
the runner^ which gradually slopes from the front of the
furnace to the farther end of the cast house. The runner
passes down the middle of the cast house, and from the
runner the sand gradually slopes towards each side of the
house. A skimmer is arranged a short distance in front
of the furnace. This is formed by leaving an opening at
the bottom of the runner, but closing it over above, so that
the iron can pass through, but the slag floating on the iron
is held back. When considerable slag has accumulated,
the sand at one side is broken through near the top and
the slag is allowed to run off through a trough provided
for the purpose.

40 MANUFACTURE OF IRON §32

On one side of the runner, a series of parallel troughs pass
from the runner to the side of the cast hoUse, and, connect-
ing with these troughs, a series of molds are made in the
sand. The sand is broken through, allowing the iron to run
into the troughs, and from these it passes into the molds,
where it is allowed to cool. The iron that cools in the molds
is known as pigs and that which cools in the troughs is
known as sows. After the iron has cooled sufficiently, so
that sudden cooling will not hurt it, water is sprinkled over
it. The pigs are then broken loose from the sows, the sows
are broken into suitable lengths, and the whole is loaded on
cars. The sand is then worked over and molds made on the
opposite side of the house to receive the next dkst.

REACTIONS IN THE FURNACE

46, There are many reactions in the blast furnace; a
number of these are known to occur regularly, but we will
probably never be able to learn positively all that take place
under different circumstances. It is not our purpose here
to point out all possible reactions, but merely to give the
most important of those known to occur. Just how some of
the important reactions take place is not definitely known,
but the exact method is of greater theoretical interest than
practical importance. As there are two currents in a blast
furnace — a gaseous current passing up and a solid current
passing down — the reactions may be viewed from two stand-
points. We will look at the matter in both ways.

46. Cliangres in the Gaseous Current. — As the blast
of hot air enters the furnace at the tuyeres, it comes in con-
tact with the fuel of the charge, heated to incandescence at
this point, and the oxygen of the blast unites with the car-
bon of the fuel, forming carbon monoxide. Whether carbon
monoxide or carbon dioxide is formed first is a disputed
point, but this is a matter of little moment, for if carbon
dioxide is formed first, it is immediately changed to carbon

§32 MANUFACTURE OF IRON 41

monoxide on coming in contact with more fuel, according to

the equation

C0, + C-2C0

It seems most probable, however, that carbon monoxide
is the first product. This is the principal reducing agent of
the blast furnace. In passing upwards, it meets the highly
heated ore, which has been rendered porous by the heat of
the upper part of the furnace, and unites with the oxygen
of the ore, according to the equation

dCO + Fe^O, = dCO^ + %Fe

thus accomplishing the reduction of the ore with the pro-
duction of metallic iron. The carbon dioxide formed when
the ore is reduced immediately comes in contact with more
incandescent fuel, and carbon monoxide is again formed.
This acts on a second portion of ore, and these reactions
continue until a point in the furnace is reached at which
the temperature is too low to induce these reactions. The
gases then pass up and out through the downcomer with-
out further change. While the reaction just given is usu-
ally mentioned as the principal reducing reaction, it is by
no means the only reaction, for carbon monoxide never com-
pletely reduces iron, and at high temperatures, if the iron is
in the spongy form, it acts as an oxidizing agent to a certain
extent, and carbon dioxide oxidizes it quite energetically.

The carbon monoxide also reduces the ore according to
two other equations, viz. :

Fefi^ +C0 = 2FeO + CO^
and FeO + CO = Fe+ CO,

It acts as an oxidizing agent according to the equations

Fe+CO = FeO+C
%FeO +C0 = Fe^O, + C
and 2/v + ^CO = Fe^O, + dC

The carbon dioxide acts on the hot, spongy iron and par-
tially reduced ore according to the equations

42 MANUFACTURE OF IRON §32

%Fe + 3(7(9. = Fefi^ + ^CO
and %FeO + CO^ = /v.O, + CC>

We thus see that oxidation and reduction are taking place
side by side, the reactions depending on the temperature and
the proportions of the elements entering into the reactions
at different points in the furnace ; but as the reducing ten-
dencies are greatly in the majority, the iron is finally com-
pletely reduced, though not by carbon monoxide alone, as
we shall presently see.

The gases that pass out through the downcomer are com-
posed of carbon monoxide and carbon dioxide (from the
oxygen of the air and ore, and the carbon of the fuel and
carbon dioxide from the limestone), nitrogen from the air,
moisture from the stock, and small quantities of volatile
matter from the fuel, hydrogen, and other constituents
not frequently determined. Though the nitrogen takes
no active part in the reduction, it serves a useful pur-
pose. It is hot when it enters the furnace and becomes
intensely heated in the vicinity of the tuyeres; then as
it passes up through the stock, it gives up much of
its heat, driving off moisture and preparing the ore for
reduction.

47. Reduction of tlie Ore. — In studying the action of
the gases, we have seen that carbon monoxide reduces the
ore and sets free metallic iron, and this is the most econom-
ical method of reduction ; but the ore is never completely
reduced in this way, and reduction is always taking place in
two other ways at the same time. In both of these cases,
the reduction is accomplished directly by the carbon of the
fuel. In the one case, each atom of carbon unites directly
with 2 atoms of oxygen of the ore and escapes as carbon
dioxide. In the other, each atom of carbon takes 1 atom of
oxygen from the ore and escapes as carbon monoxide. A
little study will show that the reduction by carbon monoxide
is the most economical, while the last method mentioned is
the most expensive.

§32 MANUFACTURE OF IRON 43

One pound of iron in the form of hematite ore is com-
bined with 4 pound of oxygen, and the heat absorbed in
reducing this is 1,886 calories, no matter how the reduction
is accomplished. One pound of carbon burning to carbon
monoxide develops 2,481 calories, and ^ pound of carbon will
be required to reduce the pound of iron ; hence, in reducing
by means of carbon monoxide, |^ X 2,481 = 797 calories are
developed when the carbon is oxidized to carbon monoxide
and /j X 5,699 = 1,799, or a total of 1,799 + 797 = 2,696,
calories are developed by burning the carbon, while 1,886
calories are absorbed in reducing the ore, leaving a surplus
of 2,696 — 1,886 = 710 calories to heat the furnace.

When the ore is reduced by 1 atom of carbon taking 2 atoms
of oxygen, forming carbon dioxide directly, only one-half as
much carbon is required for the reduction, or -^ pound will
reduce 1 pound of iron. In this case, we have -^ X 8,080
= 1,298 calories developed, but only half the carbon is con-
sumed. The other half will burn to carbon monoxide,
developing ^X 2,481 =399 calories. Adding this to the
heat devoloped in reducing the ore, we have 1,298 + 399
= 1,697 calories developed; but 1,886 calories are used in
reducing the ore ; hence, by this method we have a deficit of
189 calories, which must be made up by additional fuel
before any of the fuel can be used to heat the furnace.
When the ore is reduced by carbon with the formation
of carbon monoxide, we have -^-g X 2,481 = 797 calories
developed and 1,886 calories consumed; hence, there is
a deficit of 1,089 calories in this case to be made up by
extra fuel.

Of course the reduction is never accomplished by any
one of these methods alone, but all are going on side by side.
The carbon dioxide formed when reduction takes place by
the second method will be reduced to carbon monoxide, and
this will reduce a further quantity of ore if the conditions
are favorable ; and the carbon monoxide formed during the
reduction by the third method may reduce a second quan-
tity of ore, if it comes in contact with it under proper
conditions. When the furnace is working badly, the

44 MANUFACTURE OF IRON §32

gases may escape after the first reaction, and then, of
course, there is a loss. The ratio of CO to CO^ in the
escaping gases will give an idea of how the furnace is
working.

48. Other Reactions. — ^When the iron is reduced, it
forms a spongy mass that, in contact with the incandescent
fuel, absorbs carbon. It is now thought that much of the
carbon taken up by the iron is the finely divided carbon
deposited when carbon monoxide is decomposed by the
spongy iron.

The carbon absorbed makes the iron more fusible, and
it melts and trickles down to the hearth. At the same
time, silicon, phosphorus, and manganese are reduced and
unite with the iron. Practically all the phosphorus in the
stock goes into the iron. The amount of silicon and car-
bon depends largely on the temperature. With a hot fur-
nace, the amount of combined carbon will usually be low,
but the iron will contain much graphite. A hot furnace also
tends to produce an iron containing a large amount of sili-
con, but this will depend on the burden. If an excess of
lime is present, the silicon will mostly unite with this and
leave the iron rather low in this element. Under ordinary
conditions, most of the manganese passe,** into the iron,
though some goes into the slag. As a rule, the hotter the
furnace in which the iron is made, the less sulphur the iron
will contain, but this also depends on other conditions. If
an excess of lime is present, much of the sulphur will unite
with this, even though the furnace may not be very hot,
while if silica largely predominates in the burden, consid-
erable sulphur will pass into the iron, even though the fur-
nace be very hot. While these changes are taking place, the
basic material of the limestone, ore, and coke ash, consisting
principally of alumina, lime, and magnesia, unites with the
silica of the ore and fuel, forming a fusible slag, which melts
and trickles down to the hearth, where, on account of its
lighter specific gravity, it forms a layer above the molten
iron in the hearth.

§3a MANUFACTURE OP IRON 45

SliAGS

49. Composition of Slag:. — Slag, or cinder, as it is
frequently called around the furnace, is usually considered
as a double silicate of lime and alumina, but part of the lime
is usually replaced by magnesia. All of the constituents
vary with the kind of stock used, and when running with
the same kind of stock, the proportions of fuel, flux, and ore
will be varied from time to time, thus changing the com-
position of the slag in order to produce certain results.

The slag from a furnace using charcoal as fuel will usu-
ally contain less alumina than that from a coke furnace, for
charcoal contains little or no alumina, and slags from char-
coal furnaces are usually quite silicious.

The slags from coke furnaces are less silicious and usu-
ally contain more alumina, as the ash of the coke contains
considerable alumina. Lime and magnesia may replace
each other through quite a wide range without materially
affecting the charax:ter of the slag, and alumina may appar-
ently replace either to a limited extent. Alumina is a weak
base, and in some cases may even act as an acid to a cer-
tain extent, thus rendering the slag more acid than the
analysis would indicate. It is thought that this is most
likely to occur when the slag contains considerable alumina
and magnesia. Some of the magnesia may then unite with
alumina, forming magnesium aluminate (spinel), and neither
the alumina nor the magnesia thus combined takes any part
in fluxing silica. It should be stated that some metallurgists
think that alumina always plays the part of an acid in a
furnace slag. It seems more probable that it ordinarily acts
as a weak base, but is known to act as an acid sometimes.

A rule frequently given for the slag of a coke furnace is
that the sum of the silica and alumina should amount to
about 49 per cent, of the slag, and in blowing in a furnace,
the burden is frequently calculated so that the slag shall
contain 60 per cent, of silica and alumina. The following is
the analysis of an ordinary slag produced at a coke furnace,
and as a rule a slag having very nearly this composition is

46 MANUFACTURE OP IRON §32

sought at coke furnaces. It should be remembered, how-
ever, that the lime and magnesia may replace each other.

SiO^ = 34.26

FeO^ .32

CtfC> = 43.57

MgO= 4.23

CaS= 3.28

As a rule, the determination of silica and alumina is all
that is required as a guide to the practical running of the
furnace. If the slag is very acid, more of the iron enters the
slag as ferrous oxide, forming a fusible, scouring slag that
rapidly attacks the lining.

Some of the phosphorus passes into the slag with the
iron, but much of the sulphur passes into the iron and white
iron usually results. On the other hand, if the slag is basic,
nearly all the sulphur passes into the slag ; but if very basic,
the slag is very difficult to fuse and there is likely to be
trouble in removing it from the furnace. If the furnace is
kept hot enough to fuse the slag readily, an extravagant
fuel consumption is necessary, and the iron is so overheated
that it is likely to be of poor quality, and, as most of the
silica passes into the slag, the iron will be low in silicon.
From what has been said, it will be apparent that the com-
position of the slag must be governed by the desired
composition and quality of the iron produced.

50. Fusibility of Slagrs. — Ordinary furnace slag, as we
have seen, is composed of silica, alumina, lime, and magne-
sia ; hence, each of the constituents, when alone, is infusible
at the highest temperature obtained in the blast furnace,
and any one of the bases combined with silica would give a
slag that would be very difficult to fuse; but when all three
bases are present in the proper proportion, a slag is formed
that fuses at a comparatively moderate temperature. An
acid slag containing considerable iron and manganese fuses
readily and is very liquid when fused. A strongly basic
slag is difficult to fuse and is thick and sluggish. A small

§32 MANUFACTURE OF IRON 47

amount of magnesia is thought to increase the fusibility of
slag, but a large amount is thought to raise its fusing point.
There is much conflicting evidence in regard to the fusibility
of slags, and the most that can be stated positively may be
summed up in the three rules for the fusibility of silicates.

1. Silicates of fusible bases, such as the alkalies, are more
fusible the more base they contain.

2. Infusible bases form silicates that obtain their maxi-
mum fusibility for a certain proportion, while any other
proportion diminishes their fusibility.

3. In the case of the less fusible silicates, a multiple
silicate is more fusible than a simple one.

51. Practical Handling: of Slag^s. — When the slag is
to be flushed off, the solid slag that has chilled at the inner
part of the cinder notch is broken through, and the molten
slag, which is forced out by the blast, runs down a trough
to the cinder car, which stands on a track low enough for
the slag to run into it. The slag is hauled in this car to
the cinder dump, where it is emptied. As the slag is running
down the trough, portions of it are dipped out by means of
a ladle and poured into a cast-iron mold, where it is allowed
to solidify. This will take but a moment, and as soon as
solid may be placed on a stone or the ground to cool, while
the mold is used to receive a second portion of slag. When
cold, these test pieces are broken and examined, and to the
experienced eye they show a great deal in regard to the con-
dition of the furnace. If the slag is basic, the interior of
the piece will be gray or white, and when strongly basic, the
white or gray may extend to the surface. A normal slag is
usually gray in the center and dark towards the surface. If
the slag is black and glassy when broken, and thin pieces
that break off are transparent or translucent, it indicates
that the slag is acid in character. If the slag is brownish
and dull in color, it indicates that the furnace is working
cold. From these test pieces, a sample is selected and taken
to the laboratory fpr analysis.

48 MANUFACTURE OF IRON §32

Various attempts have been made to utilize slag, but
without much success up to the present. Some of it is used
as railway ballast, and some as a road-making material.
Sometimes the slag is led into a tank of water, and a jet of
water is caused to impinge upon it as it flows in. This
causes it to swell up, forming a brittle, spongy mass, much
of which floats on the water and may be raked off. When
removed it crumbles up like coarse sand, in which form it is
used somewhat as a building material and in making foot-
paths. A jet of steam is sometimes blown into the running
slag, thus blowing it out like spun glass. In this form, it is
known as slag wool, and is used to a certain extent as a non-
conducting covering for steam pipes. Its use in this form,
however, is very limited. A method of utilization that
seems to have met with considerable success in some places
is to mix the granulated slag with lime, making cement of it.

CAIiCUIiATION OP BURDENS

62. When using a new mixture of stock, it is necessary
to calculate the proportions of the constituents necessary to
produce the desired slag, or to refer to one of the tables pre-
pared for this purpose. If a table is used, it should be
remembered that these results are obtained by making cal-
culations with stock having a certain composition, and are
therefore only approximations with stock of different com-
position, and should always be verified. Having once
started, the charge of fuel is fixed, and this remains the
same, while the ore and limestone are varied as circum-
stances may require. What burden a furnace will carry
with a given weight of fuel depends on the fuel itself and on
the ore to be smelted. Methods have been given for the
calculation of the weight of ore to be charged with a given
weight of fuel, from the analysis of the ore aijd fuel, but as
so much depends on the physical structure of the fuel, the
only way to determine this accurately is by an actual trial
in the furnace. After having fixed the weights of fuel and
ore to be used, the weight of limestone to be added may be

§32 MANUFACTURE OF IRON 49

found by referring to a table or by calculation, and then the
correctness of this weight may be checked by the method
of verification to be given ; or, we may assume a weight of
stone and verify it, and if the result is not what we wish,
we can change the weight of stone as the result of the cal-
culation indicates to be necessary. We prefer to calculate
the weight of stone and then to verify this as follows:

Let us assume that the coke in the charge is fixed at
8,425 pounds. We must allow for 5 per cent, loss due to
moisture, dust, etc. ; hence, this would give us 8,000 pounds
as a basis of calculation. Then, let us assume that this
charge will carry 14,500 pounds of ore, and that the analyses
of the ore, coke, and limestone are as follows:

Ore Coke Ash

Iron = 55.0jl^ SiO^ = 5.83^

StO^ = 10.3j[^ ^',^, = 3.08^

Al^O^= 2.6^ CaO = .28^

CaO = 2.8^ Mg-0 = .11}<
MgO =1.9^

Limestone

StO^ = d^

Al,0,= in

CaO =60}<

MgO= %i

Now let us assume that we wish to produce an iron con-
taining 2 per cent, of silicon and a slag containing about
34 per cent, of silica.

In discussing slag, it was stated that some metallurgists
regard the alumina in a slag as an acid, and that it was
sometimes given as a rule that a slag should contain 49 or
50 per cent, of silica and alumina. For the purpose of cal-
culation, let us assume that the alumina acts as an acid,
and that we wish to produce a slag in which the sum of the
silica and alumina will amount to 50 per cent. Then,
arranging the constituents according to their acid or basic
character, we have

60 MANUFACTURE OF IRON §32

Ore

Basic Acid

CaO = 2.8j^ SiO^ = 10. 3j^

Total, Zt^ 12.95<

Coke

Basic AcicL

CaO = .28^ SiO^ = 5.SS^

MgO = .Hi Al^O^ == 3.08^

Total, .39^ 8.91^

Limestone

Basic Acid

CaO z=z 50^ SiO^ = d^

MgO - 2j^ Al^O^ = Ij^

Total, 52^ 4^

14,500 X 4.70^ = 681.5 pounds basic material in ore.
14,500 X 12.90^ = 1,870.5 pounds acid material in ore.

8,000 X .39^ = 31.2 pounds basic material in coke.

8,000 X 8.91^ = 712.8 pounds acid material in coke.

Arranging these, we have

Basic Acid

Ore 681.5 1b. 1,870.5 1b.

Coke 31.2 lb. 712.8 lb.

Total 712.7 lb. 2,583.3 lb.

Thus, we find that in one charge of ore and coke we have
712.7 pounds of basic material and 2,583.3 pounds of acid
matter; but we want 2 per cent, of silicon in the iron, and
the ore contains 55 per cent, of iron ; hence, there will be
14,500 X 55j^ = 7,976 pounds of iron made from each charge.
As silica is nearly one-half silicon, it takes about 4 per cent,
of silica to yield 2 per cent, of silicon; hence, we have 7,975
X 4j^ = 319 pounds of silica to supply silicon to the iron.
This must, of course, be subtracted from the total acid

§32 MANUFACTURE OF IRON 51

material, leaving 2,683.3 — 319=2,264.3 pounds of acid
matter to go into the slag. There are also 712.7 pounds of
basic matter present, which will unite with an equal weight
of acid matter to form a slag containing 50 per cent, of acid
matter, leaving 2,264.3 — 712.7 = 1,661.6 pounds of acid
material to unite with the basic material of the limestone.
The 4 per cent, of acid matter in the limestone will unite
with an equal amount of basic matter, leaving 62 — 4
= 48 per cent, of basic matter available for fluxing the acid
matter of the ore and coke. As there are 1,551.6 pounds of
acid matter to be fluxed by the stone, 1,551.6 pounds of basic
matter of the stone will be required ; and as the stone only
contains 48 per cent, of available basic material, we will
need 1,551.6 -^ .48 = 3,233 pounds of limestone.

This is, of course, only the amount of limestone necessary
to produce a slag containing 50 per cent, of silica and
alumina. We have not calculated the amount of silica
alone, but with ordinary stock, when the silica and alumina
compose 50 per cent, of the slag, the percentage of silica
will be about right. This, however, should be verified as
follows :

Considering the alumina as a base, from the analyses of
the constituents already given, we have

Ore

Basic Acid

Al^O, = 2.^^ SiO^ = 10. 3j^

CaO = %,%^
MgO = \.9^i

Total, 7.3ji^ 10.3;^

Coke

Basic Acid

Alfi^ = 3.08^ SiO^ = 5.835^

CaO = .28^
MgO = .llj^

Total, 3.47^ 5.83j^

52 MANUFACTURE OF IRON §32

Limestone
Basic Acid

CaO = 50^
MgO = 2^

Total, 53^ 3j<

Arranging these, we have

14,500 X 7.30j^ = 1,068.50 pounds basic material in ore.
14,500 X 10.30^ = 1,493.50 pounds silica in ore.

8,000 X 3.47^ = 277.60 pounds basic material in coke.

8,000 X 5.83^ = 466.40 pounds silica in coke.

3,233 X 53.00^ = 1713.49 pounds basic material in lime-
stone.

3,233 X 3.00j^ = 96.99 pounds silica in limestone.

Basic Acid

Ore 1,058.50 lb. 1,493.50 lb.

Coke 277. 60 lb. 466. 40 lb.

Limestone 1,713.49 lb. 96.99 lb.

Total 3,049.59 lb. 2,056.89 lb.

Thus, we have 2,056.89 pounds of silica in each charge.
Subtracting the 319 pounds of silica that goes into the iron
as silicon, we have 2,056.89 — 319 = 1,737.89 pounds of
silica to go into the slag, together with 3,049.59 pounds of
basic material. Then, dividing the weight of silica by the
total weight of slag-forming material, we have 1,737.89
-f- 4,787.48 = 36.51 per cent, silica.

This slag would be all right in blowing in a furnace, but
would be rather acid for ordinary running, so we will need
to add more limestone. From the analysis of the stock, we
would judge that it would require between 600 and
700 pounds of limestone to bring the slag down to 34 per
cent, of silica, so we will try 3,900 pounds of limestone next.
Then, we would have

3,900 X 53.00^ = 2,067 pounds basic material in limestone.
3,900 X 3.00^ =117 pounds silica in limestone.

§32 MANUFACTURE OF IRON 53

Taking the figures previously obtained for ore and coke,
we have

Basic Acid

Ore 1,068.5 lb. 1,493.5 lb.

Coke 277.6 lb. 466.4 lb.

Limestone 2,067.0 lb. 117.0 lb.

Total 3,403.1 lb. 2,076.9 lb.

Subtracting the 319 pounds of silica that goes into the
iron, we have 2,076.9 — 319 = 1,757.9 pounds of silica that
goes into the slag, and dividing this by the total slag-form-
ing material, we have 1,757.9 -r- 5,161 = 34.06 per cent, of
silica in the slag.

CliASSIFICATION OP IRON

63. Iron is usually classified as Bessemer, basic, mill,
malleable, charcoal, and foundry iron, depending on the
purpose for which it is to be used ; and the purpose for which
it is to be used will govern its composition.

Bessemer iron is for use in the manufacture of Bessemer
steel, and as practically all the phosphorus and sulphur in
the iron remain in the steel, the percentage of these elements
must be low. By Bessemer iron is usually meant an iron
containing less than .1 per cent, of phosphorus and less
than .05 per cent, of sulphur.

Basic iron is to be used in the basic process of steel manu-
facture. The iron should contain as little silicon as possible,
as this will attack the basic linings. For the same reason
the surface of the iron should be free from sand. By this proc-
ess, the phosphorus is largely removed, and, consequently,
basic iron may contain considerably more phosphorus than
would be permissible in Bessemer iron.

Mill iron is for use in the puddling mill, for the manu-
facture of wrought iron. It should contain a low percentage
of silicon, and the iron made when the furnace is working
badly -on f oimdry iron is sometimes used for this purpose.

64 MANUFACTURE OF IRON §32

Malleable iron is used for making malleable castings. It
usually contains more phosphorus than Bessemer iron and
less than foundry iron, and the percentages of silicon and
graphitic carbon are low.

Charcoal iron is simply iron made in a furnace using
charcoal as fuel. It is generally used as a foundry iron for
special purposes.

Foundry iron is used in making castings, by melting it and
pouring it into molds ; hence, for this purpose, an iron that
will readily fill the mold and will not shrink on cooling is
desired. The other properties of the iron will depend on the
character of the castings to be made.

64. Grading^ by Fracture. — When foundry iron is taken
from the pig bed, it is loaded on cars, and these are placed
on a trestle beside which triangular or wedge-shaped blocks
of cast iron are fastened on top of strong supports. The
iron is then thrown from the cars on to these blocks, breaking
each pig in half, and the broken iron is piled according to
the appearance of the broken surface.

No. 1 iron is dark gray in color and the grain is large and
even. The iron that is a little lighter in color or having
smaller grain, or in which the size of the grain is not quite
so even, is called No. 2 x. If the grarn is a little too small or
uneven, or the iron is a little too light colored for 2 x, it is
graded as No. 2 plain. No. 3 iron is close-grained, and is
usually lighter colored than the other grades. This holds with
iron containing less than 3 per cent, of silicon. If the iron
contains over 3 per cent, of silicon, the portion of it having a
fracture that would be graded as 1, 2 x, and 2 plain is graded
as Scotch iron, and that having a close grain is sold as high
silicon 3.

66« Grading by Analysis. — The fracture of the iron
indicates, to a certain extent, the kind of castings for which
it is adapted, and formerly foundrymen depended on the
fracture entirely, but at present a chemist is employed at
most of the large foundries, and the composition of the iron
is taken into account. Recently, it has been suggested that

§32 MANUFACTURE OP IRON «5

the fracture be disregarded and that the foundry mixtures
be made to depend on the composition of the iron entirely.
Those who have tried this plan report excellent results.
While there has as yet been no general agreement as to
what shall constitute the several grades, at least one large
concern has printed specifications to govern its purchases.
These specifications will probably be modified in time, but
they serve well as a starting point. They are as follows:

Foundry No. 1

Silicon must not be less than 2.60fl

Sulphur must not exceed OSjt

Phosphorus should not exceed 60jt

Manganese should not exceed 50fl

Foundry No. 2

Silicon must not be less than 1 . 95^

Sulphur must not exceed 04}<

Phospiiorus should not exceed 70^

Manganese should not exceed lOjt

Foundry No. 3

Silicon must not be less than 1 . 35j^

Sulphur must not exceed 05j^

Phosphorus should not exceed SOjf

Manganese should not exceed 90}<

If this method of grading should be generally adopted, both
upper and lower limits will probably be adopted by general
consent. At present, most foundry iron is purchased by a
combination of the two methods. The purchaser orders a
certain grade of iron (graded by fracture) having a certain
composition. •

EliEMENTS CONTAINED IN IRON

56. Carbon. — Carbon occurs in iron in at least two con-
ditions— graphitic and combined carbon. Its affinity for iron
varies with the temperature and the percentage of other ele-
ments in the iron. In ordinary pig iron, the percentage of

56 MANUFACTURE OF IRON §32

carbon will seldom exceed 4.6 per cent., but high manga-
nese iron and chrome iron may contain as much as 7 per
cent., and, it is claimed, even more than this.

Carbon has a remarkable power of distributing itself
through iron, tending to become uniformly distributed not
only through one piece, but through several pieces in contact
when hot. When the iron is in the molten condition in the
furnace, all the carbon is thought to be in the combined
state or dissolved in the iron, but as the iron cools graphite
separates throughout the iron. The formation of graphite
in iron is favored by high percentages of total carbon and
silicon, and is opposed by the presence of sulphur and man-
ganese. It is generally said that graphite has little direct
influence on the character of iron beyond lowering its tensile
strength, but it appears to be the general experience of fur-
nacemen that an iron containing a high percentage of
graphite is darker in color and softer than one containing
less of this form of carbon. Combined carbon increases the
tensile strength and hardness of iron, but diminishes its
ductility.

To obtain an iron with a large amount of graphite, it is
necessary to have a high temperature in the hearth of the
furnace and a strongly reducing atmosphere, in order that
much carbon may be taken up by the iron. To obtain these
conditions, the temperature of the blast should be high and
the burden should be light ; that is, the proportion of fuel to
ore should be large. These conditions favor a high percent-
age of total carbon and also a high percentage of silicon,
which causes much of the carbon to take the graphitic form
on cooling. It is generally stated that a basic slag, on
account of its refractory character, promotes the formation
of graphite; but as a basic slag reduces the percentage of
silicon, this statement can scarcely be considered as an
established fact. An aluminous slag is probably advanta-
geous.

To obtain an iron high in combined carbon, we may run
the furnace with a heavy burden and an acid slag. This
will keep the hearth at a lower temperature, so that little

§32 MANUFACTURE OF IRON 57

silicon is reduced and the acid slag allows considerable sul-
phur to enter the iron, and sulphur tends to increase the
percentage of combined carbon, as does also manganese.

67. Silicon. — Silicon readily unites with iron, forming
iron silicide, which dissolves in the iron. Iron containing as
much as 20 per cent, of silicon can be made in the blast fur-
nace, but when more than about 6 per cent, of silicon is
present, the product is known as ferrosilicon. Unlike car-
bon, silicon seldom occurs in iron in the uncombined state.
It cannot be reduced from its combinations by either carbon
or iron alone, but is reduced by the combined action of the
two. Silicon diminishes the power of iron to combine with
carbon, so that in the presence of a very high percentage
of silicon the total carbon will be lower than if less were
present, but, as we have seen, it increases the graphite by
lowering the percentage of combined carbon. This property
of changing combined to graphitic carbon is probably its
most valuable one in relation to iron, for in this way it
makes the iron softer and tends to lessen the shrinkage of
castings, though the silicon itself would tend to increase this
shrinkage. It tends to prevent the formation of blowholes
in iron, by increasing the solubility of the enclosed gases,
and makes the iron more fusible.

To produce an iron high in silicon, a high temperature in
the hearth is necessary, hence, a light burden and strongly
heated blast are generally employed. A slightly acid slag
containing considerable alumina to make it refractory are
advantageous. We should not attempt to reduce more than
25 or 30 per cent, of the silica of the stock, for extravagant
fuel consumption is necessary to accomplish this, and it is
better to use more silicious ores, if we wish to produce iron
containing more silicon than will be furnished to the iron by
this percentage.

68. Phosphorus. — Phosphorus combines with iron in
all proportions up to 26 per cent. It is found in iron as
phosphide of iron, or possibly as the phosphides of iron and

58 MANUFACTURE OF IRON §32

manganese, dissolved in the iron. It tends to prevent blow-
holes, makes the metal more fluid, and is thought to
prevent shrinkage on cooling, so that the metal fills the
mold more perfectly; hence, a moderate amount of it is
desirable in foundry iron. On the other hand, it makes the
iron brittle, giving it a tendency to break under suddenly
applied loads. Iron containing much of this element is
treacherous, as it will sometimes bear a heavy load, and
again may be broken easily. It is said to lower the point of
saturation of iron for carbon. Phosphides of iron and man-
ganese are the only compounds of this element formed in
the furnace, and as both of these are soluble in iron, practi-
cally all the phosphorus in the stock goes into the iron and
only a very small portion enters the slag. Consequently, if we
know how much phosphorus our stock contains, we can tell
almost exactly how much the iron will contain, and if a lower
percentage is desired, we must use stock containing less of it.

59. Mangranese. — Manganese alloys with iron in all pro-
portions. It increases the tensile strength and fluidity of
iron and makes it harder and less fusible. It has a stronger
affinity for carbon, sulphur, and oxygen than has iron, and,
consequently, it will remove oxygen and sulphur from iron
and produce an iron with a high percentage of carbon. It
prevents the formation of blowholes, by preventing boiling
while cooling, and by reducing and removing oxide and
silicate of iron. It unites with sulphur, forming a compound
insoluble in iron, and thus, to a large extent, removes this
element from iron. It is also thought to counteract the effect
of other impurities in the iron, tending to prevent red short-
ness, but does not prevent cold shortness due to phosphorus.
It tends to make sound castings, by preventing blowholes
and removing oxides and silicates.

It raises the saturation point of iron for carbon and
prevents the separation of this element as graphite on cool-
ing; hence, it tends to produce an iron high in combined
carbon. Iron containing much manganese is usually low in
graphite and high in combined carbon. On this account.

§32 MANUFACTURE OF IRON 59

much manganese is thought to hinder the production of
high-grade foundry iron. Distinct names are given to the
alloys of iron with considerable manganese. Alloys contain-
ing from about 10 to 25 per cent, of manganese are known
as spiegeleisen^ and those containing from 25 to 90 per cent,
of manganese are known as ferromanganese. These
alloys usually contain but little graphite and silicon and
much combined carbon. Manganese is very difficult to
reduce ; hence, in making these alloys, a very hot blast and
light burden are necessary. Dolomite is generally used as
flux, as the magnesia makes a more difficultly fusible slag
than does lime, and the slag should be basic. The fact that
the slag is basic would account for the low percentage of
silicon in the product; but in addition to this, manganese
probably has a tendency to lower the percentage of silicon.
Even though a light burden and hot blast be used and a
basic slag with a high percentage of magnesia be employed,
the manganese will not all be reduced, especially if the ore
contains much silica, but some of it will pass into the slag,
giving it a green color, and making it fluid and corrosive.

60, Sulpliur. — Sulphur combines with iron in all pro-
portions up to 53 per cent. It forms a number of sulphides
of iron, but in pig iron it usually occurs as FeS dissolved
in the metal.

Much sulphur makes the iron hard and brittle and pre-
vents the separation of carbon as graphite ; hence, iron con-
taining a high percentage of sulphur also, as a rule, contains
much combined carbon. A high percentage of sulphur
causes blowholes, but makes the iron more fusible. An iron
containing much sulphur is usually low in silicon and vice
versa. This may be due to the fact that high sulphur iron
is usually made when the furnace is not hot enough to pro-
duce iron high in silicon, but these two elements appear to
be antagonistic in iron. Sulphur may be expelled from iron
in the furnace by a number of agents, as basic slags, man-
ganese, and calcium. The effect of magnesium in elimina-
ting this element is a matter of dispute. Some authors say

60 MANUFACTURE OF IRON §32

that a magnesian slag will remove sulphur equally as well as
a calcareous one, while others say sulphur will not unite
with magnesium at all in the furnace. Manganese is very
efficient in removing it. So strong is its affinity for sulphur,
that if ferrous sulphide and manganese are fused together,
the manganese will take the sulphur from the iron, uniting
with it to form a slag. The sulphur in the ore is more
easily removed when it exists in the form of sulphate than
in sulphide, and that in the fuel is more easily expelled than
that in the ore. When a very low percentage of sulphur is
desired, a basic slag should be used, and if this does not give
a sufficiently low percentage, poorer ore should be used in
the mixture and more limestone should be added to produce
a larger volume of slag.

61. Arsenic. — Arsenic will probably combine with iron
in nearly all proportions, but is not one of the usual con-
stituents of iron. Coke and limestone are usually free from
this element, and few ores contain it. It is contained in
some ores, however, and when present, some of it volatilizes
in the furnace and some passes into the iron, where it seems
to act much like sulphur. It appears to lower the satura-
tion point of iron for carbon, to give it a white fracture, to
make it red short and brittle at high temperatures, and if
much is present, it makes the iron cold short.

63. Titanium. — Titanium is not one of the usual con-
stituents of iron, but small quantities of it are always likely
to be found in iron smelted from magnetite Its effect on
iron is not known, but it does not appear to be injurious.
Magnetite ores are always likely to contain this element and
should be examined for it. Ores containing it may be used
to a limited extent, but the slag formed is much more
refractory than with ordinary ores, and they are always
likely to cause trouble in the furnace.

63. Copper. — Copper occasionally occurs in iron ores,
and small quantities of it are sometimes found in iron.

§32 MANUFACTURE OF IRON 61

especially in iron containing considerable manganese. A
small quntity of it is not objectionable in foundry iron, but
its presence should be avoided in iron to be used in the man-
ufacture of steel, as it is said to make steel red short.

PRACnCAIi SUGGESTIONS

64. Blast. — A certain amount of air is needed to burn
the fuel of the furnace, and as a hot blast is almost univer-
sally used, it carries heat into the furnace, and conse-
quently, up to a certain point, the more blast that is used,
the hotter will the furnace become. But as the tempera-
ture of the hearth is approximately 3,000° F. and the tem-
perature of the entering blast is from 900° to 1,500° F.,
after we pass this point the blast has a cooling effect, and
necessitates the use of njore fuel to keep up the tempera-
ture.

At the present time the amount of blast used exceeds that
necessary to produce the maximum temperature, thus secur-
ing a larger output at the expense of increased fuel consump-
tion. Taking all things into consideration, it is more
economical to drive in this way, and the rate of driving gives
lis the most convenient and immediate method of regulating
the temperature of the furnace.

If the furnace becomes too cold, we may heat it up by
reducing the number of revolutions made by the blowing
engines in a minute. This will tend to raise the tempera-
ture of the furnace at once, but at the same time the stoves
should be changed at frequent intervals, in jorder to main-
tain a uniform high temperature of blast. If this does not
heat the furnace sufficiently, or if it becomes cold again on
returning to the previous rate of driving, it indicates that
too heavy a burden is being carried, and the burden should
be reduced. When this new burden reaches the hearth, the
old rate of driving may be resumed.

If the furnace is too hot, it may be cooled by increasing
the number of revolutions of the blowing engines, and if

ea MANUFACTURE OP IRON §32

this does not cool it sufficiently, a little cold air may be passed
in with the blast, thus reducing the temperature of the
blast. If the furnace becomes too hot again when the nor-
mal rate of driving is resumed, it indicates that too light a
burden is being carried, and the burden should be increased.
As we have seen, it requires a high temperature to reduce
silicon ; hence, the amount of silicon in the iron can be
largely controlled by the blast. If the furnace is running
cold, the percentage of silicon in the iron will be low. By
increasing the temperature of the blast and reducing the
rate of driving, the furnace will be heated so that more sili-
con will be reduced, and the slower rate of driving will
leave the stock longer in the heated portion of the furnace,
so that there is more time for the reduction of this element
and consequently the next iron made will contain more of it.
The same result may be obtained by using a lighter burden,
but in this case no change in the percentage of silicon can
be obtained until the stock charged in the new proportion
has had time to reach the hearth. The rate of driving may
be decreased and the burden reduced at the same time, and
when the new burden reaches the hearth the old rate of
driving may be resumed. On the other hand, if the iron is
too silicious, this may be corrected at once by harder dri-
ving and, if needs be, by using cold air to reduce the temper-
ature of the blast. At the same time, a heavier burden
may be put on, and when this reaches the hearth, the orig-
inal rate of driving may be resumed. There are many
things that affect the temperature in the furnace, and as
changing the rate of driving is the quickest and handiest
way of regulating the temperature, the blowing engines are
seldom run at the same rate for 24 consecutive hours.

65. To Detect Xieaking: Tuyeres. — The tuyeres are
subjected to very destructive influences, and sooner or later
will wear out and leak. The water passing into the hearth
of the furnace chills it and injures the quality of the iron.
There are several methods of detecting leaks. AVTien a
tuyere is leaking, the blast will frequently force the water

§32 MANUFACTURE OP IRON 63

along its surface to the outside and the joint of the wall
will become damp. A larger volume of gas than usual and
its peculiar appearance and odor are good indications that
water is entering the furnace. If in doubt about a tuyere
leaking, the water may be slackened for a moment so that
it has less pressure than the blast. If upon turning it on
again it discharges white, it shows that the blast has
entered it, and, consequently, that there must be a leak. If
one end of a stick is held between the teeth and the other
end placed against the tuyere pipe and the ears stopped,
any flow of water into the furnace can be detected. If a
cold steel bar is run into the furnace through the tuyere
while the blast is off, it will show moisture when withdrawn,
if the tuyere is leaking. A brass or copper tube filled with
water is better than a steel bar for this purpose. It should
be kept in a cool place.

66. Tuyeres Taking: Blast Irregularly. — Blast may be
prevented from entering tuyeres by obstructions, and more
blast consequently enters the open tuyeres. This makes the
furnace work faster on the side of the open tuyeres, caus-
ing slipping of the stock and an intense local heat. When
the tuyeres are taking the blast irregularly, so that some are
dark and others bright, we would ordinarily think that the
bright tuyeres were taking blast freely while the dark ones
were closed, but this may not be the case. If a large vol-
ume of blast is entering a tuyere when the heat is low, it
may chill cinder on the nozzle, giving a dark tuyere, while
the opposite tuyere, which is taking less blast, is bright. If
the blowpipe is now tested, the dark tuyere will show its
blowpipe much hotter than that of the bright tuyere, which
is not receiving enough blast to chill the cinder, but merely
to cause an intense local heat. This tuyere requires picking
with a rod, to get an opening well into the hearth that the
blast may enter. A dark tuyere and hot blowpipe show
that the tuyere is taking blast freely. A bright tuyere and
hot blowpipe indicate the same. A bright tuyere and
cold blowpipe show but little blast and poor penetration.

64 MANUFACTURE OF IRON §32

A dark tuyere and cold blowpipe shows that the tuyere is
closed. This may be caused by a piece of scaffold. A rod
should be used to make an opening through to fresh coke,
that the blast may enter. If this fails, a cartridge may be
used.

Neglect of tuyeres causes increase of pressure, imeven
settling of stock, scaffolding, poor iron, and the burning
out of tuyeres. Large hearths require more attention to
the tuyeres than do small ones, in order to secure even dis-
tribution of blast.

67. Scaffolds. — If on account of slow driving or from
any other cause the stock above the fusion limit is highly
heated, it becomes pasty and in passing down the boshes it
is pressed against the walls and adheres to them. This
hard ring on the walls holds up the stock above it, forming
what is called a scaffold, while the stock passes down the
center of the furnace. This causes irregular working and a
small output of poor iron. The heat gradually works up
through the stock above the ring on the boshes, and some-
times reduces much of the iron and forms a pasty mass of
fuel and limestone, which is partially cemented together
with slag. Consequently, the longer a scaffold remains in a
furnace, the worse it is likely to become. When scaffolds
are first formed, they may frequently be removed by char-
ging blanks of fuel and scrap iron. If this fails, they may
often be removed by charging fuel and then drawing back
the tuyeres, cutting them away with a large volume of
blast, allowing the ring or scaffold to come down in front of
the tuyeres. If there is now sufficient fuel below the scaf-
fold from the blanks previously charged, it will be melted,
and the furnace will work regularly in a short time. If
there were not sufficient fuel at the tuyeres, the scaffold
coming down would chill the furnace.

When the tuyeres are taking blast irregularly and some-
times from other causes, a lump will form on one side of
the furnace, holding up the stock above it, while the other
• side remains clear. This is known as a side scaffold. It

§32 MANUFACTURE OF IRON 66

causes the stock to settle faster on one side than on the
other, and thus may be detected by watching the way the
stock settles at the top. It also frequently makes one side
of the furnace shell hot. These scaffolds may frequently be
removed by charging scrap on the side of the scaffold, and
some material, like fine ore or anthracite, which does not
take the blast freely on the other side. The quickest way
to get rid of a side scaffold is to have holes in the side and
bosh walls, where they are likely to occur, and as soon as
they form to crack them loose with giant powder. The pre-
caution must be taken, however, to have extra fuel in the
hearth at the time, to melt them up and prevent chilling the
furnace.

68. Hsmfging and Slipping:. — Sometimes, especially
when much fine ore is being used, the stock in the upper
part of the furnace sticks to the walls, while that below con-
tinues to settle, leaving a space between the two portions of
stock. This is known as hanging. If we continue to drive
as usual, this stock may hang for some time and then sud-
denly slip down, causing a violent disturbance in the fur-
nace, known as a slip or explosion. As soon as it is found
that the stock is hanging, an attempt should be made to
cause it to settle, so that it will not have so far to slip when
it does come down. This is best accomplished by turning
off the blast, which, of course, has a tendency to hold the
stock up, for a few moments at intervals of about 10 min-
utes," until the stock comes down. If there is much iron in
the hearth when the stock is caused to come down by throw-
ing off the blast, there is always danger of its being forced
up around the tuyeres and destroying them. There is also
danger of slag being forced up in the same way ; hence, the
slag should be tapped off before throwing off the blast, and
if anywhere near casting time, the iron should also be
tapped off.

69. Trouble Wltli Iron Notcli. — Carelessness in closing
the iron notch after casting may often cause trouble. The
clay should not be forced into the tapping hole as soon as the

MANUFACTURE OF STEEL

(PART 1)

INTRODUCTORY

!• Definition of Steel. — While at first thought it seems
to be a simple matter to define steel properly, the more famil-
iar one becomes with the subject, the more perplexing is it to
write a concise definition that will apply to the wide range
of steels produced, or even to the greater part of them.
Before the introduction of the modern methods of manufac-
ture, the distinction between steel and wrought iron was
sharp and well marked, and steel could then be defined as
**any alloy of iron with carbon that would take a temper on
quenching." Wrought iron does not sensibly harden on
sudden cooling in water from a red heat. Modern methods
of manufacture, however, have produced a metal that largely
partakes of the nature of wrought iron, yet is made by the
same processes, that give a metal that hardens on quench-
ing. For this reason such a classification as the above
would now throw out the greater amount, or at least a very
large tonnage, of the material classed and accepted by the
metallurgical and commercial world as steel. The Bessemer
converter and open-hearth furnace early showed an adapta-
bility to produce a soft metal having great strength, elas-
ticity, and ductility, capable of displacing wrought iron, and,
for most purposes, far superior to it. Anything that follows
is not offered as a thoroughly comprehensive definition of
steel, as none can be offered that is not easily assailable and
its inapplicability shown from some standpoint.

§33

For notice of copjrright, see page immediately following the title p«g«.

2 MANUFACTURE OF STEEL § 33

Steel may be defined as a metal produced by the complete
fusion of materials in a bath, the necessary properties being
given, after conversion, by additions of carbon or carbon
alloys. Wrought iron may be defined as a metal produced by
the partial fusion, or bringing to a pasty condition, of mate-
rials on a hearth.

** Blister," or ** cementation," steel, made by soaking bars
of iron, at or above a red heat, in charcoal or carbon, would
seem to be a notable exception; but as this is mainly an
intermediate product for remelting in crucibles, and its pro-
duction being of little importance, it will be disregarded in
this treatment of the subject.

2, The question of the proper classification of steels is
one to which much attention has been given in the past, an
international committee at one time having been selected
from the metallurgical and technical societies of the princi-
pal steel-producing countries to adopt a universal classifica-
cation. While much good came of their work, and strenuous
efforts were made to adopt their classification, neither metal-
lurgically nor commercially was it ever generally used.

Many theories have been advanced as to what steel is.
One that is held by many practical metallurgists is that the
ideal steel is an alloy of pure iron and carbon only, all other
elements being regarded as impurities. From this point of
view, all grades of steel can be produced by simply varying
the amount of carbon; but as impurities are necessarily
present, all steels contain varying, and usually very small,
amounts of sulphur, phosphorus, silicon, metallic oxides,
and gases, which require other additions for their neutrali-
zation or elimination. Again, special alloys are required for
giving steels characteristic qualities for particular purposes;
such are the nickel, tungsten, chrome, manganese, and
molybdenum steel.

3. History, — Steel was probably first made in Asia or
Northern Africa by the Chaldeans, Egyptians, or other
early civilizations, by methods probably more like the cru-
cible process than any we have record of today. In fact, a

§ 33 MANUFACTURE OF STEEL 3

very limited amount of steel, but of most excellent quality,
is still made in India (called Indian or Wootz steel) by
reducing very pure ores, mixed with chopped wood, in clay
crucibles heated by a charcoal fire blown by goatskin bel-
lows. From this steel, the celebrated Indian sword blades
were made, than which no finer tool steel has ever been
produced.

Our interest in present methods of manufacture dates
from the invention of the crucible process, in 1740, by Ben-
jamin Huntsman, of Sheffield, England, a clockmaker dis-
satisfied with the quality of cementation steel in clock
springs. This remained practically the only method of
production for over a century, when in 1855 the Bessemer
process was invented by Henry Bessemer and the regenera-
tive open-hearth furnace by the Siemens, Messrs. Charles
William and Frederick, in 1861. Not until these processes,
especially the Bessemer, had produced large quantities of
steel much cheaper than the crucible, did steel begin to
supplant wrought iron to any great extent and thereby
inaugurate the **age of steel." It is this vast tonnage of
cheap steel that has rendered possible the' wonderful indus-
trial development of the world in railroad and ship building,
the varied lines of engineering and construction affecting
every nation of the world and the condition of each
individual.

4, Processes of Mannflax^ture. — There are only three
processes for the manufacture of steel : The crucible^ the
oldest of present methods; the Bessemer ; and the open-
hearth. The last two were developed almost simultane-
ously. The Bessemer was first perfected, and for the first
35 years, or up to about 1890, led the open-hearth, both as
to tonnage produced and in the perfection of methods and
appliances — both metallurgical and mechanical. While the
Bessemer process still produces the greater tonnage, this
is the only direction in which it can claim superiority over
the open-hearth. In the order of their metallurgical and
commercial importance today the processes rank: first, the

4 MANUFACTURE OF STEEL § 33

open-hearth; second, the Bessemer; and third, the crucible.
They will be treated in this order.

While the crucible process is of the least consequence, it
holds the most distinctive field metallurgically, and one
from which the others seem unlikely to crowd it. Given
th^ same composition, it is well established that crucible
steel is superior to either of the others, but owing to the
much higher cost of production, its use is now restricted
mainly to the making of high-grade tools, certain mining
drills, parts of intricate machines, and, in general, where
the first cost of the steel can be ignored.

The open-hearth process has a larger field it can claim as
its own than the Bessemer. Open-hearth steel is now used
for the better grades of plate steel, forgings, car axles, and
structural steel. The basic open-hearth process is used
where an extra soft, pure steel is required, as in plates,
sheets, rods, wires, etc. Bessemer steel is used for rails,
nails, tin plate, light axles, in fact, for those articles where
cheapness is desired. It is, however, being rapidly replaced
by steel produced by the basic open-hearth process. The
basic process, by cheaper production than was possible in
the acid open-hearth, makes this a formidable rival of the
Bessemer and seems practically assured to largely supplant
it in the next few years. Owing to lower cost of produc-
tion, the Bessemer process heTd undisputed sway for years
in all lines using a large tonnage of steel. The open-hearth
gradually demonstrated its superior fitness for special lines.
While both the crucible and open-hearth processes have dis-
tinctive fields, held from the cheaper metal by the superior
quality of their product, the Bessemer has no field the open-
hearth cannot fill, and only by lower cost does it still pro-
duce the greater tonnage. Practically all rails are as yet
made of Bessemer metal, also most of the ** billets and
slabs " for merchant bar, tin plate, sheets, nails, and light
axles; some ship and tank plate, etc.

Some of the reasons for the cheaper production by the
one or the other process, for their special fields and uses,
will be treated under their respective heads.

§ 33 MANUFACTURE OF STEEL 6

THE OPEI^^-HEARTH PROCESS

5. Historical. — Steel was first made by the open-hearth
process in England, in 1862, in the regenerative furnace of
the Siemens brothers, which was patented in 1861, but
which was developed and perfected by Charles William Sie-
mens, who is better known by his title, Sir William. This
was not the first attempt to make steel on an open hearth,
however, many previous experiments having been made,
notably those by Josiah Marshall Heath, in 1845. But it
was only with the Siemens apparatus, which gives the high
temperature necessary, together with an almost perfect
control of heat conditions, that success was possible.
Siemens efforts were originally directed to producing steel
by the reduction of iron ore in a bath of pig iron without
the use of scrap; the ore, by its reduction, furnished the
oxygen for oxidizing the carbon, silicon, and manganese of
the pig metal.

About 1864, the Messrs. Martin, French steel makers,
made steel by melting pig iron and scrap in the Siemens
furnace, and patented the process. In France and some
parts of Europe it is still known as the Martin-Siemens^ or
Martin^ process^ but in Great Britain and America as the
Siemens-Martin^ or more generally in recent years merely
as the open-hearth process. The above ternis are frequently
indiscriminately used, but it should be clearly understood that
the Martins never laid claim to the regenerative furnace, but
only to the pig-and-scrap process worked in the Siemens
furnace, for which entire credit is due them, while the fur-
nace is wholly a Siemens production. It is correct to speak
of the Siemens-Martin process (pig and scrap), but only of
the Siemens furnace.

Th% pig-and-ore (sometimes incorrectly called the direct)
and the pig-and-scrap processes, have for years been used in
combination. In the past few years, owing to the rapid
expansion of the open-hearth industry, also in improve-
ments in rolling-mill methods, the amount of available
scrap has been so reduced that metallurgists have been

6 MANUFACTURE OF STEEL § 33

forced to use the pig-and-ore process, the two most success-
ful methods of which will be considered in detail.

6. Open-Heartli Furnace. — The open-hearth furnace
consists of a rectangular hearth approximately twice as long
as it is wide; the term operi simply signifies that the hearth
is to be so constructed at both ends. This form is one of the
oldest of metallurgical furnaces, but the regenerative prin-
ciple of the Messrs. Siemens constitutes its originality and
value. By regeneration is meant the giving up of the waste
heat of the escaping gases and the temporary storing of it
in such a way that the air for combustion is always pre-
heated, or regenerated; producer gas is always preheated,
but natural gas is not. By this means a very much higher
temperature is obtained than is otherwise possible, as
well as great fuel economy. The hearth is connected by
means of the ports and vertical flues with chambers,
called regenerators, placed at a lower level either directly
under the hearth, or, preferably, set back so as to be less
readily choked up by the fine dust, soot, etc. carried over
by the current of escaping gases and by slag and metal,
which sometimes cut through the bottom or sides of the
hearth.

Four chambers — two at each end, one for gas and one for
air — are built to each furnace. Each chamber is connected
at the bottom with suitable flues, which have valves control-
ling the gas and air supply, so arranged that the currents of
gas and air can be reversed at regular intervals, usually of
15 minutes; the incoming supply travels through the regen-
erators, through which the waste gases escaped during the
previous interval. This constant reversal of the direction of
gas and air, and of the ends at which they are introduced
into the furnace, is kept up during the melting. Theoret-
ically, the only limit to the temperature attainable in a
regenerative furnace is the point of dissociation of hydrogen
and oxygen, about 2,500° C. (4,532° F.). This point, how-
ever, can never even be approximated practically, owing to
the limit set by the inability of the refractory materials to

§ 33 MANUFACTURE OF STEEL 7

withstand such a temperature and the rapid loss of heat by
radiation at high temperatures.

7. Construction of the Open- Hearth Furnace. — Two

types of furnaces are in general use: The fixed, or station-
ary, furnace and the tilting, or rolling, furnace. In both
types the furnace proper, or melting chamber, is the same —
rectangular in section and connected with regenerators, as
has been explained. It is covered with an arched roof of
9 or 12 inches of the best grade of silica brick; the side
walls are also made of the same material, usually 9 inches
thick. Silica bricks expand, about ^ inch to the foot in
heating to a working temperature, and to partially allow for
this, they are never laid close. Further allowance for this
expansion is made in the construction by a system of tie-
rods having turnbuckles, or nuts, so that they can be
lengthened as the furnace heats and the bricks expand.

The hearth is built in a pan of heavy riveted plate steel
carried on beams supported on a solid block of concrete and
brick, or on heavy foundation walls, or piers, so that the
weight of the furnace and charge is not carried on the regen-
erator arches, if these are under the furnace. Other beams
are set perpendicularly along the sides and ends, their ends
connected beneath and above the furnace by tie-rods. Steel
rails were formerly used for this purpose, but they have been
supplanted by I beams. These rails or I beams, called btuk-
stays^ are connected by means of tie-rods at top and bottom
and serve to keep the furnace sufficiently rigid. Without
these the structure would not stand the strains due to the
weight of the charge and the expansion and contraction as
the temperature changes.

8. Roof. — For many years the roof of the furnace was
thrown from the side walls ; that is, the weight of the roof
was carried by the walls, just as the weight of any arch is
carried on the walls from which it springs. This construc-
tion was objectionable for many reasons, and caused serious
trouble when the side walls of the furnace **cut out," as

8 MANUFACTURE OF STEEL § 33

frequently happens, while the rest of it is good. In such
cases it was practically impossible to repair the walls, and the
weight of the roof soon caused them to fall. The side
thrust on the walls also caused their distortion, and as they
wore down this became more serious. The present method
of construction obviates these objections by carrying the
roof on heavy channels, in which the skew back (the beveled
brick on which the arch starts) is placed so that almost the
entire weight of the roof is carried by the two channels,
thus relieving the walls. In this way, when the. side walls
fall in or are partially burned out, they may readily be
renewed or patched (which is frequently done) without dis-
turbing the roof. Or at the end of a run, if the roof is in
good condition, other repairs necessary may be made and
the old roof used for the next run. This i& not general prac-
tice, as it is customary at most plants to put on a new roof
for each run of a furnace.

While the construction of open-hearth furnaces varies
greatly, a 4:0-ton stationary furnace is shown in Figs. 1
and 2. Fig. 1 is a longitudinal section of the right-hand
half through the center and a side elevation of the left-
hand half. Fig. 2 (a) is a cross-section on the line A B oi
Fig. 1. These figures illustrate a common form and show
the principle of all open-hearth construction.

9. Siemens Reprenerator. — The air and gas chambers
are built of the same length and height and extend at
right angles to the furnace hearth. The air chambers are
about one and one-third times the width of the gas chambers,
a greater volume of air being required than of gas. Both
chambers contain checkerwcfrk of brick, usually the best
quality firebrick or silica brick, so laid as to expose a
large surface to the gases. Sometimes the construction is
such as to give a number of small horizontal flues in each
chamber, but more generally the brick are staggered in or
baffled, alternate courses being placed over the parallel
passage below, in both horizontal and transverse courses.
This is done to distribute the current of waste gases and

lifQyalSme^

StnffQug/. smcm

/#*^4/#/ /7/» Br/ek A

n^Qym/.Rim Brick C

i^slQuai. Fif Brick A ^8888881

i'if' Qua/. F/r9 Brick B
C/t0iywf nocr

<.''

j^L

Pig. 1

g 33 MANUFACTURE OF STEEL 9

bring them more intimately in contact with the )^.rick sur-
faces of the checkers, assuring a better absorption of iieat and,

in turn, a more thorough reabsorption of this stored heat by
the incoming gas and air when the currents are reversed.

10 MANUFACTURE OF STEEL § 33

In Fig. 1, a and g show, respectively, the air and gas cham-
bers on one end containing the brick checkerwork (the oppo-
site end is exactly the same, but the chambers are not shown) ;
from the chambers the vertical flues, or uptakes u^^ w^, lead
to the ports /„> A> ^^^ ^^^ being carried above the gas;
on the other side of the furnace, at the same end, are cor-
responding flues leading from the opposite end of the air and
gas chambers, so that on each end of the furnace there are
two air uptakes and ports and two gas uptakes and ports.
Frequently, one large gas uptake leads from the middle of
the gas chamber, terminating in one port between the air
ports. The simplest way to understand the relations and
functions of the chambers and ports is to consider them as
parts of one huge gas burner; the supply of gas and air
comes from the respective chambers and is conducted by the
tubes (uptakes and ports) to where they can mix and com-
bustion take place, i. e., in the melting chamber, where the
heat is wanted. The bottom, or hearth, is shown at b. The
roof is made of 9-inch or 12-inch silica brick, 9-inch in this
case. The flues f under the checkerwork connect them with
the valves and draft stack. The *'slag pockets" $ extend
under a part or all of the furnace ; they are a continuation
downwards of the uptakes, their purpose being to catch any
slag, brick, etc. and keep it out of the chambers.

The left-hand section of Fig. 1 shows the elevation from
the top of the furnace to the bottom of the chambers ; also
the beams, tie-rods, etc. for supporting and strengthen-
ing the structure. The hydraulic, or pneumatic, cylinders r, c
are for raising and lowering the furnace doors by means of
chains passing over the sheaves s\ s\ /; they are controlled
by valves conveniently placed on the charging floor.

Fig. 2 (a) is a cross-section through the hearth of the
furnace on the line A B oi Fig. 1. Fig. 2 {d) is an end
elevation of the furnace. Fig. 3 {a) shows a horizontal
section of the flues on the line CD of Fig. 1. It shows the
flues a and g with their connections a' and g' to the air
and gas valves v^ and Vg^ for one end of the furnace and to
the stack. The dampers in the chamber and stack flues are

1 83 MANUFACTURE OF STEEL 11

shown at d. Fig. 3 {b) is a sectiQn on the line G H, showing
the air and gas reversing valves v, and v^, and the regula-
ting valves V for each.

Tracing the course of the gas and air, we have the gas and
air entering through the valves, thence through the flues to
the chambers shown in Figs. 3 and 3; the uptakes and ports

12 MANUFACTURE OF STEEL § 33

now conduct it to the melting chamber for combustion ; the
waste gases passing out at the opposite end through the
chambers and flues to the stack. At the end of 15 minutes
the reversing valves are thrown and the gas and air pass
in the opposite direction. This reversal at regular intervals
of the currents is continued throughout the working of the
furnace. There is no mixing of the gas and air until they
are brought together at one end of the hearth for combus-
tion. It sometimes happens that a communication is estab-
lished between them previous to this by the cutting through
or wearing away of a division wall, when premature com-
bustion takes place — the gas always being hot enough to
burn readily after passing a very short distance through the
chamber. In such a case the hearth is robbed of just that
amount of heat besides the serious injury to a part of the
furnace not designed or capable of withstanding the tem-
perature produced.

As the gas and air first enter the hot regenerators, the
latter are cooled, as no heat is produced until the gas and
air meet in combustion in the furnace beyond. The flame
here begins to heat the furnace and also the regenerators
at the other end, as the waste gases pass through on their
way to the stack. When the chambers, on the end at which
the gas and air enter, are cooled somewhat and those on the
opposite end correspondingly heated, the reversing valves
are thrown so that the gas and air travel in the opposite
direction. By this means the regenerators are constantly
becoming hotter, so that the heat produced by the combus-
tion of the gases is a continually increasing quantity — a
thermal arithmetical progression. The hotter the gas and
air (within limits here attainable), the higher is the tem-
perature they produce on combustion. The regular reversal
of the gase?, which by going through the regenerators and
becoming constantly hotter, produces a constant increment
of temperature in the melting chamber. This is so great
that without the careful regulation of gas and air a furnace
would ** melt itself down " in a short time. This is especially
true in an empty furnace, or towards the end of a heat when

§ 33 MANUFACTURE OF STEEL 13

the stock is all melted and the metal hot. Such a condition
can scarcely come about during the melting-down stage, as
the bath is then rapidly absorbing heat.

10. Ports, — The ports are the openings or passages
through which the gas and air are led into the furnace
hearth, combustion taking place at their mouths. They are
connected with the regenerative chambers by what are com-
monly termed the uptakes. There is no part of the furnace
requiring greater care in design and construction, for on
their size, proportion, and arrangement, proper combustion
depends more than on any other point.

There are usually two gas and two air ports at each end
of the furnace. This is varied by two gas and three air or
one gas and two air, the air in any arrangement always being
on the outside and above the gas, because the air is the
heavier, and by having it on top of the gas as the two spread
out and mix at the port ends, combustion takes place, and
the flame is thrown towards the bath. By this means not
only is the heat kept on the stock or bath, but the cutting
action, aside from, or in connection with, the temperature
produced, has much to do with the melting down of the
stock. An important point is to keep the flame away from
the roof, as the latter may cut out or be melted down with
improper port design. Another reason for having the air
on top is to avoid the oxidation that would be produced by
a layer of hot air striking the stock or bath. The air and
gas should meet about 2 feet above the metal, according to
some authorities 5 feet, but this will bring it too near the
roof in the ordinary furnace. If they meet much less than
two feet above the metal, combustion can hardly begin freely
before it is checked by striking the stock or bath; if much
more, the most intense temperature is so high above the
bath that, the toof and sides suffer.

The pitch that the ports are given is an important matter;
if too flat, the flame is not brought down sufficiently on the
metal and combustion is too high up in the melting chamber.
The tendency in such a case is for the brickwork to receive

MANUFACTURE OF STEEL

§ 33 MANUFACTURE OF STEEL 16

the maximum temperature rather than the metal; if too
steep, the flame is brought down upon the stock before com-
bustion is completed when the full heat value of the gas is
not developed ; besides, there is a tendency for the heat to
concentrate in one place and not be properly distributed
over the hearth.

11. Wellman Rolling^ Furnace. — Fig. 4 shows two
Wellman rolling, or tilting, furnaces — one in the normal or
melting position, the other in position to pour steel. The
furnace consists of a strongly framed steel casing approxi-
mately rectangular in section, inside which the brick lining
is built up. On the under side are fixed two curved rockers
that roll and are supported by strong steel bracings; when
tilting to pour off, the furnace moves forwards on these
rockers. The movement is accomplished by two hydraulic
cylinders r, placed underneath and the upper ends of their
piston rods attached to the pouring side. To tilt the fur-
nace, waten is admitted to the top end of the cylinder when
the piston is pulled down. In case of accident or failure of
the water pressure, the furnace returns by its own weight
to the level position. -

The sides and ends of the furnace consist of steel structural
work tied together and stiffened with plates, angles, and
tie-rods across the top of the furnace body, as shown. Each
end of the furnace has openings at g^ for the passage of the
gas and at a for air, around which is fitted a cast-iron water-
cooled ring ^, which fits into a corresponding ring on the port
when the furnace is upright. The ports differ from those
of the ordinary furnace in that they are built inside a
strongly framed steel structure d and are separate from the
body of the furnace, being carried on four flanged wheels.
The uptakes from the regenerators are carried to about the
level of the furnace bottom, and across the top of each is
laid a short track on which the wheels of the port structure
rest. Two cast-iron water troughs e extend around the
upper part of the uptakes, and on the under side of the port
openings are rings that project into the water troughs, thus

16 MANUFACTURE OF STEEL § 33

forming a water-sealed joint between the movable port and
fixed uptake, preventing the leakage of gas and air in pass-
ing in or out of the furnace. As mentioned above, the joint
between the furnace body and ports is made by water-
cooled rings in each, so that both the vertical and horizontal
joints of the ports allow practically no leakage.

When about to pour, each port is drawn back to avoid the
friction between them and the furnace ends. The channels
across the top of the port structures act as bails by which
they can be picked up from their track by an overhead
crane, set aside, and a fresh pair placed in position.

The regenerative chambers are arranged in the same
general way as in the ordinary furnace, but are always
placed back of the furnace under the charging platform.
The valves for reversing and controlling the gas and air are
similar to those of the fixed furnace. The lining in acid
furnaces is silica brick, both on the sides and on the roof;
in basic rolling furnaces, the magnesite bricks are carried in
the back wall so as to be above the slag when the furnace is
being poured, as the basic slag would flux with any silica
brick. The tapping hole is so arranged as to be always
above the level of the bath when melting; it is fitted with
a heavy flanged-steel casting riveted to the furnace body;
holes in the outer flange serve to readily attach either the
forehearth or pouring spout, if a ladle is used for casting.

1J5. Foreheartli. — This part of the furnace may be
described as a special ladle attached to the front of the
tapping hole, and is a special feature of the Wellman rolling
furnace, and was developed by Mr. S. T. Wellman. It
allows the steel to be poured directly into the molds without
the use of a ladle. It is a box-shaped casting shown at/*,
with a flanged opening on one side corresponding to the
tapping hole to which it is bolted. It is brick-lined and is
provided with two pouring holes and stoppers. When the
furnace is tilted, the metal flows into the forehearth and is
thence tapped into the ingot molds, on cars, which are
pushed along under the forehearth to be filled. Each car,

§ 33 MANUFACTURE OF STEEL 17

or bogie^ usually carries two molds, which are placed the
same distance apart as the pouring holes, so that two molds
can be filled at once, thus facilitating the casting oper-
ation. The forehearth, while performing the function of
the casting ladle to a certain extent, differs from it in
that it does not become a reservoir for any considerable
amount of metal, but acts more as a passage for the metal
from the furnace to the ladle. A pouring spout may be
readily substituted for it and the steel run into the ladle as
in the ordinary practice.

13. Advantagres of the Rolling: Furnace. — Rolling
furnaces have come into extended use in the past few years,
and their future seems to be assured. Some of the reasons
for this are the following:

No trouble results, nor is time lost in taking care of the
tapping hole, as this is always above the level of the bath,
and must be stopped simply to exclude air, hence no time
is lost tapping out.

It permits the ready removal of slag. This becomes of
greater consequence as impure irons, producing large
amounts of slag in the basic process, are used.

The partially reduced metal is easily transferred from one
furnace to another, and the slag is got rid of at the same
time.

In pouring, as the joints with the ports are broken, the
gas must be shut off; this at first seems a disadvantage, but
is the reverse, as the cold air admitted at the ends chills
the surface of the slag without affecting the temperature of
the metal appreciably, and prevents boiling and violent
action while pouring.

Holes form in the bottoms of all furnaces even with the
most careful attention. In this event, fixed furnaces must
be bailed out with rabbles, and this can frequently be only
partially done, so that metal is left to be absorbed by the
bottom, which becomes more or less soaked with it and
oxide of iron, thus very greatly reducing its power to with-
stand the action of slag and metal. In the tilting furnace,

18 MANUFACTURE OF STEEL § 33

all the metal and slag can be removed after each heat, leav-
ing the bottom dry ; a considerable saving of metal results
from this, as well as better preservation of the bottom.

It offers special advantages for the Talbot and Bertrand-
Thiel processes, which are described later.

14. Capacity of Open-Heartli Furnaces. — The early
furnaces had a capacity of from 3 to 5 tons, but they were
gradually increased with the development of the process,
construction, and means for readily handling the large
amounts of stock and product. So far as the successful
working of the furnace is concerned, there is practically no
limit to the size of the furnace, but in taking care of the
product, obstacles are met. The largest furnaces that have
yet been constructed, in which the entire melt is withdrawn
at once, take a charge of 120,000 pounds and yield about
50 gross tons of ingots. These furnaces have a melting
chamber about 33 feet long and 14 feet wide. This will
probably remain the standard size for large furnaces for
some time. There are a number of reasons for this, not
metallurgical and engineering alone, but economical as well.
The prompt handling of a mass of 50 tons of molten steel
within the allowable time and under the conditions of
pouring is an engineering feat of such magnitude that it
has been accomplished only within the last few years.

Present conditions demand that all the heat possible be
saved, and for this reason steel from the. furnaces must be
put through the rolling mills as soon as possible, in order
to take the least amount of reheating. With much larger
heats than the above coming at one time, some of it will
take a large amount of reheating before it can be put
through the mills.

Another objection is the time required for pouring or cast-
ing (sometimes cMed deeming). Molten steel is really a deli-
cate fluid and the limits of temperature within which it can
be handled to produce good steel, or to avoid spoiling good
steel, are not very wide. It is here that the skill and train-
ing of the steel maker count, in particular that of the

§ 33 MANUFACTURE OF STEEL 19

melter or blower, as the case may be. If so large a heat is
made that it cannot be poured rapidly, it must either be
too hot at the beginning to make good steel, in order to get
all of it out of the ladle and avoid a ** skull'* or ** chilled
heat,'* or of the proper temperature at the beginning, with
the result that it will be too cold at the end.

16. Gas and Air Valves. — These have been briefly
spoken of, but a fuller description is demanded by their
importance in furnace operations. The admission of gas
and air is regulated by a simple form of throttle valve.
Besides, there are reversing valves for changing the direc-
tion of gas and air. Both sets of valves are controlled from
the melting floor by levers or by a hand wheel and screw,
connected by suitable rods or chains. Many forms have
been patented, but the ideal valve has not yet been invented,
as all give more or less trouble in furnace operations.
Among these troubles are the cracking or warping of
the seat or the box due to the uneven temperature to
which they are subjected. A deposit of soot and tar in the
gas valve requires cleaning, or leaks ensue from failure of
the valve to close tight. Only two of the many types will
be described.

In any type of reversing valve there is a box, or outer
casing, made of cast iron or steel plate within which the
reversing valve proper works. Attached to the top of this
box, or casing, is the regulating valve that controls the gas
or air ; in the former case, it is connected to the gas main,
and in the latter, it opens to the air for its supply. The valve
box sits over three openings; the one in the center connect-
ing with the flue that goes to the stack ; the ones on either
end with flues going to the regenerators on corresponding
ends of the furnace. The opening to the stack flue and
chamber on one end are always connected, the position of
the valve directing the gas or air through the flue on the
opposite end to the corresponding regenerator, passing on
up to the hearth, where combustion takes place. The waste
gases are led through the regenerators on opposite ends and

ao MANUFACTURE OF STEEL §33

back to the valves, the-position of the latter directing the
waste gases (admitted into the valve box from the flue
beneath), downwards into the flue that is connected to the
draft stack.

16. Siemens Valve. — The Siemens, or butterfly, valve
is the oldest form of reversing valve and is still largely
used. It is the simplest and, in many respects, the best
type yet devised. Fig. 5 shows this valve in sectiori. It
consists of the outer casing, or box a, described above, and
the elliptical tongue, or butterfly ^, which is the valve

proper. This is suspended by arms through its center rest
ing in the sides of the box, one end protruding and connect-
ing with a lever for reversing. The elliptical ends pi the
butterfly fit in corresponding sections of the valve box so as
to make, as nearly as possible, gas-tight and air-tight joints.
The objections to this type of valve are: (1) The warping and
cracking of the cast-iron tongue and box so that gas leaks
through to the stack, as the pull to the latter is stronger
than the pressure of the gas to the furnace. (3) It is
exposed to the hot producer gas on one side and the waste
gases on the other, so that cracking and warping frequently

g 33 MANUFACTURE OF STEEL 21

occur, causing delays
in changing and in-
creasing the cost of
repairs. (3) A de-
posit of soot and tar
around the joints pre-
vents the valve
closing tightly and
allows gas to leak.

While the waste
gases pass to the stack
for the most part at
600° to 800° F., they
occasionally escape at
red heat, when the
valve suffers. Water
cooling of both valve
and box has been
" tried, but with little
£ success. The advan-
tages of this valve are
simplicity and cheap-
ness, so that even if
requiring frequent re-
pairs, they can be
made quickly and at
a comparatively small
cost.

17. Forter Valve.

The troubles with the
Siemens valve have
led to an almost end-
less number of valves
being designed to
avoid its defects.
Water cooling of the
parts in contact with

22 MANUFACTURE OF STEEL § 33

the hot gases is the essential feature of most, and a
water seal of many. The **Forter" is perhaps the most
perfect of this type, the general arrangement of which
is shown in Fig. 6. The base plate, or trough casting ^, is
made of cast iron and holds about 2^ inches of water. It
has three openings having flanges the height of the out-
side flange, corresponding with those in the brickwork to
connect with the stack flue b in the middle and the regen-
erator flues c to the furnace on either end. Two of the
openings are covered by a movable plate-steel or cast-iron
hood A, connecting one or the other of the regenerator flues
with the stack flue. This hood performs the office and cor-
responds to the ** butterfly" in the Siemens valve. It is
carried on arms d that lift it out of the water seal in revers-
ing, describing an arc of a circle, moving so as to connect
the opposite regenerator flue and stack flue and is dropped
into the water seal in its changed position. This movement
is accomplished by an outside lever connected to a shaft
controlling the inside lifting arms ; this shaft and the bot-
tom edges of the hood are under water when seated, thus
making a gas- or air-tight water seal. Running water is
supplied to the base plate at one end to keep the seal cold
and replenish the loss by evaporation, the overflow being
carried off at the other end.

18. Cut-Off Valves. — In addition to the regulating and
reversing valves, each furnace using producer gas has a cut-
off valve so placed that the gas can be completely shut off
from any furnace without interfering with any other, in case
of removal or repair of the other valves or when a furnace
between others connected to the same main gas flue is out
for repairs.

The preceding description is of a furnace using producer
gas; with natural gas as a fuel, the valve arrangement is
much simplified, for, as was previously stated, this gas is not
regenerated, but fed directly to the ports from the gas line.
The gas valve then becomes an air valve also, both chambers
being used as air regenerators. The gas valve gives much

§ 33 MANUFACTURE OF STEEL ^23

less trouble than the air valve, as the air comes in hot, while
the gas serves to cool the valve and preserve its life.

19. Dampers. — The flow of the waste gases to the stack
is controlled by a damper in the stack flue, usually at the
base of the stack. Dampers should also be placed in each
flue leading from the reversing valves to the regenerators, for
while frequently omitted, the volume of waste gases passing
through the chamber determines the temperature of the gas
and air for combustion. It often happens during the run of
a furnace that one chamber becomes partially clogged up,
lessening the draft there, so that to effect an even distribu-
tion of heat to the chambers, one or the other must be
throttled; it is also sometimes advantageous to work one
chamber hotter than the other. As no tight seal is necessary
here, these dampers are merely rectangular steel or cast-iron
plates loosely fitting in the flues, controlled by a chain and
counterweight, from the charging floor.

ACID AND BASIC OPEX-IIEARTH SYSTEMS

20. General Remarks. — The open-hearth process
divides itself into the acid and basic systems. In the former
the hearth is made of acid material — silica in the form of
silica sand or silica brick ; in the latter, the hearth and such
portions of the side walls as the slag is likely to come in con-
tact with are made of basic material — magnesite or dolo-
mite— that a basic slag may be carried. The hearth is inert,
taking no part in the reactions of the process, but must be
made of a materiar to correspond to the character of the
slag produced. The slag is the active agent in effecting
purification, when this takes place, as in the basic process.
The acid is the original open-hearth method and was practi-
cally the only one worked on any important scale until 1890;
the basic is now the more important process and Is becoming
of even greater importance each year. The construction of
the furnace, with the exception of the hearth, as noted above,

24 MANUFACTURE OF STEEL 1 3

§ 33 MANUFACTURE OP STEEL 25

is identical'for either acid or basic work, the melting cham-
ber, ports, and regenerators being the same ; hence, a fur-
nace can be changed from one to the other by substituting
the one or the other lining.

21. Acid and Basic liinlng^. — The terms acid and
basic refer to the character of the lining, or more exactly
to the slag carried in the melting operation. The lining,
however, determines the slag that can be carried, as there
will be a reaction between the two, if of opposite char-
acter, until an approximately neutral* slag is reached; in
other words, the character of the slag will be changed and
the lining rapidly destroyed. Hearths of neutral material
— ^bauxite or chromite — have been unsuccessfully tried, the
idea being that a basic or acid slag could then be worked.
The terms acid and basic applied to open-hearth slags are
not absolutely strict, but relative, as an acid slag is fre-
quently basic enough to react with a sand bottom, while
a basic slag is often acid enough to react with a magnesite
bottom.

22. Wellman Cliar^cing: Machine. — Formerly all the
stock was charged in the furnace by hand. The pig and
heavy pieces of scrap were placed on a peel and guided to the
part of the hearth desired ; small and light pieces were thrown
directly in by hand, shovel, etc. This has been entirely super-
seded by the Wellman charging machine, shown in Fig. 7.
It is the invention of Mr. S. T. Wellman, who has done
more mechanically for the open-hearth process than any
one else connected with it. The first machines were oper-
ated by hydraulic or steam power, but are now operated
entirely by electricity. The machine consists of a steel
frame a carried on four wheels on tracks on the charging
floor. A movable carriage b is suspended on beams c at the
top of the machine, the beams projecting beyond the main
body of the machine, and over the track next to the furnaces
on which stand the cars with the charging boxes d. To the
front of the carriage are hung supports to which is attached

26 MANUFACTURE OF STEEL § 33

the peel, or ram (not shown in the figure), with a rectangu-
lar head for inserting into the casting on the end of the
charging box containing the pig iron, scrap, etc. Electric
motors are provided for the different motions on the track in
front of the furnaces, such as moving the carriage back and
forth to introduce the charge, and revolving the peel on its
axis to drop the stock from the box into the furnace. The
operator is carried on the movable carriage so that he has a
close view of the movements of the machine and can readily
control them by suitable levers conveniently placed. In
operation the machine picks up the box filled with stock, is
moved in front of one of the furnace doors, which is raised,
the carriage advanced inserting the box in the furnace, and
the ram revolved, dumping the stock. The operations are
now reversed and the box replaced on the narrow-gauge car.
The charging boxes are special, rectangular, steel-plate
boxes from 4 to 8 feet long and 16 to 20 inches in section,
with sides slightly flaring so that the stock will readily drop
out when overturned. The ends are of cast iron or steel,
the end next the machine is always of cast steel, as it carries
the weight of the box. The boxes are filled in the stock
yard, usually on narrow-gauge cars carrying three or
four boxes, or they are placed there by traveling cranes and
elevated or shifted to similar tracks on the charging floor,
so placed that the ends just clear the furnace buckstays.
The boxes hold from a few hundred pounds of light bulky
scrap to 4,000 pounds of pig iron, or heavy scrap. The
charging machine has done more to reduce the cost of
making steel by the open-hearth process than any single
invention or appliance; at the same time, it has taken the
hardest and hottest part of the furnace work from the men.
In a large plant, one machine charges five or six furnaces,
displacing three or four men per furnace. It is economical
even in small plants of one or two furnaces.

23. Cranes. — The electric traveling crane is the stand-
ard appliance for handling the metal and slag after tapping,
and for doing the /// work^ such as placing the molds for

g 33 MANUFACTURE OF STEEL 27

the steel, handling ingots, getting up stock, etc. The
hydraulic swing crane was formerly used and has some
advantages, such as simplicity and cheapness to install and
operate and small likelihood of getting out of order, with the
consequent delays and accidents. With this equipment one
crane was arranged to serve two furnaces by being placed
between them in a semicircular pit in front of them. In
this case the steel could only be poured into molds placed in
the pit. With the electric traveler the ladle of steel can be
picked up and carried to any part of the casting shop for
pouring or teeming. Where only top-cast ingots aie made,
the traveling crane offers the further advantage of pouring
in molds placed on cars at any convenient place within the
space covered by the crane. These can then be shifted
directly to the rolling mill, avoiding the expense and delay of
rehandling from the pit.

24. Xadle. — The steel is tapped from the furnace into a
ladle made of heavy, riveted plate steel, lined with from
4 to 6 inches of firebrick, usually two courses, the one next
to the steel shell being of a low-grade firebrick laid flat,
%\ inches thick, the inner one of a good-grade firebrick,
either laid flat or on edge, 4^ inches thick. Sometimes
only the one course is used, but this is not a safe practice for
heats of from 30 to 60 tons. The steel is always poured from
the bottom of the ladle, so as to keep the slag out of the
metal and at the same time give better control over the
casting operation. In the bottom of the ladle, near its cir-
cumference, is placed the nozzle of graphite or hard-burned
firebrick. This has a cup-shaped top, tapering to a hole
from 1 to 2 inches in diameter. The stream of metal is con-
trolled by a stopper rod, which is protected by jointed fire-
brick sleeves and carries on its lower end a graphite plug
called the stopper head. When pouring, the upper end of the
rod is connected to a slide, on the upper outside edge of the
ladle, provided with a suitable lever for opening up and
shutting off the stream of metal. Fig. 8 shows the cast-
ing side of an open-hearth plant with a 60-ton traveling

28 MANUFACTURE OF STEEL § 33

ladle crane and a 40-ton ladle in position for pouring
the heat.

Fig. 9 shows a section through an open-hearth plant. In
the figure, a is the charging machine of the low type; J,
the open-hearth furnace; /, the producer-gas main; _/', a gas

§ 33 MANUFACTURE OF STEEL

30 MANUFACTURE OF STEEL § 33

valve; //, the regenerator chamber; and ^, the stack. The
traveling crane / over the charging floor is for handling
stock, etc. At the left is shown the casting house, in which
rf is the ladle crane; r, the ladle; ^, the molds on the car; and
/, the pouring platform. The small hydraulic crane g is
used for handling the spout, setting stopper, etc.

GASEOUS FUEL USED IN OPEX-HEARTH

FURNACES

35. Introductory. — As previously stated, the operation
of the regenerative furnace depends on the gaseous fuel,
and not until the Siemens brothers developed the gas pro-
ducer was this furnace a success. A regenerative furnace,
for either melting or reheating, can be operated with natural
gas, artificial, or producer, gas, or petroleum.

NATURAL GAS

36« This is the ideal fuel, and the one generally used
where available, but it is of much less general importance
than producer gas because of its comparatively limited
geographical distribution and the probable uncertainty as
to its permanency. It was first used in the manufacture of
steel at Pittsburg in 1879, and is used principally in Western
Pennsylvania and adjacent parts of Ohio and West Virginia.
No one theory as to its origin is generally accepted, although
a number have been advanced. It is commonly associated
with oil, and is probably produced from it by distillation,
under certain conditions of temperature and pressure
within the earth, or by distillation from coal, or the two
combined. The depth of the wells varies from 1,000 to
3,000 or 4,000 feet. The pressure at the wells frequently
amounts to several hundred pounds per square inch, render-
ing it uncontrollable. In the lines, as furnished for use, a
pressure of from 6 to 10 ounces per square inch is main-
tained. It is piped considerable distances to the works,

§33

MANUFACTURE OF STEEL

occasionally as much as 200 miles. To keep up the pressure
and supply a sufficient volume of gas, special pumping
engines are used. So valuable and advantageous is its use
that a vast amount of capital is represented in developing
territory, sinking wells, and conveying the gas by means of
pipe lines to the works. While the supply is not nearly so
abundant as a few years ago, yet with the more economical
methods of handling and using it conservative experts claim
that it will last indefinitely. As fully 50 per cent, of the
open-hearth steel produced in America in 1901 was melted
with it, and perhaps one-fourth of the first reheating of the
total rolling-mill tonnage done with it, its importance in
the manufacture of steel justifies a brief account of it.

The chief advantages in its use are: (1) Higher calorific
value, with consequent increase of output ; (2) greater
purity, thus producing purer steel or allowing the use of
poorer stock ; (3) convenience and cleanliness in use.

TABIiE I

Sample

Constituent

No. I.
Per Cent.

No. 2.
Per Cent.

No. 3.
Per Cent.

No. 4.
Per Cent.

Carbon dioxide CO^ . .
Carbon monoxide CO
Oxvcren 0

.80

1 .00

1. 10

.70

3.60

72.18

20.62

.60
.80
.80
.98

5.50
65.26

26. 12

.58
.78

.98

7.92

60.70

29.03

1. 00
2 . 10

■'.^ Mm J ^-t^mM ^^ .....■■•...

Ethylene C^ff^

Ethane C,77.

Methane C/f^

Hydrogen H

Nitrofifen A^ :

.80
5.20

57.85
9.64

23.41

27. Composition of Natural Gas. — Natural gas is
essentially marsh gas, or methane CH^, with varying admix-
tures of other members of this series of hydrocarbon

39 MANUFACTURE OF STEEL § 83

gases, together with hydrogen. It usually contains from 60 to
70 per cent, of methane and 20 to 30 per cent, of hydrogen.
Table I shows the analyses of four samples, giving an idea
of its composition.

The high percentage of nitrogen in No. 4 is probably due
to- air, as natural gas seldom shows any considerable per-
centage of it. The average heating value of Pennsylvania
and Ohio natural gas is 1,007 B. T. U. (British thermal
units) per cubic foot.

28. Introduction of N'atural Gas Into tlie Furnace.

Natural gas is not regenerated (preheated), but is intro-
duced directly from the supply main into the ports of the
furnace. Regeneration was tried when the gas was first
used, but the heat of the chambers decomposed the rich
hydrocarbons* and caused a deposition of carbon in the
chambers in the form of a hard, glassy coke; it also reduced
the gas to hydrogen or lower hydrocarbons, having less
heating value than the original gas, besides losing the value
of the deposited carbon, which would be burned on a reversal
of the furnace, the products of this combustion escaping
directly to the stack instead of being utilized in the furnace
or chambers.

ARTIFICIAL. GAS

39. Under the name artificial gas, many forms and
kinds of gas have been made and used at various times,
but the only one that need be given any extended con-
sideration in connection with the manufacture of steel is
producer gas. Other artificial gases which are made by
various processes are coal gas, water gas, and oil gas,
or a gas produced by a combination of any or all these
processes. It is technically possible to use all these in
making steel, but it is not commercially possible at this
time, owing to the higher cost for producing a given
calorific effect.

§33

MANUFACTURE OF STEEL

30. Producer Gas. — The apparatus in which what is
termed the producer gas is made, is a cylindrical riveted
shell of boiler steel, lined with firebrick. The early pro-
ducers were made rectangular in section, but the circular
section was adopted as offering many advantages, and is
now wholly used. As before stated, the success of the open-
hearth furnace, or of the regenerative furnace to whatever
purpose applied, depends on the use of a gaseous fuel. The
producer may, therefore, be properly considered a part of
the furnace and its development has been simultaneous.
Producer gas may be regarded as the general fuel of
regenerative furnaces; natural gas, while superior in every
way, can be considered only as a special fuel.

31. Stemeiis Producer. — Fig. 10 shows the original
Siemens producer. It is a rectangular firebrick chamber

having one side b inclined at an angle of 45° to 60°, pro-
vided with a grate c at the bottom. The coal is fed into the
opening a at the top, making a thick bed as it falls to the
grate, through which air is admitted to the ignited fuel, and
converts a part of the carbon to carbon dio.\ide C"(?„ .which,

84 MANUFACTURE OF STEEL § 33

in passing up through the partially incandescent mass with
an insufficient air supply, is reduced to carbon monoxide CO,
by taking up an additional atom of carbon

C0,+ C=^2C0

This carbon monoxide is diluted by the inert nitrogen of
the air and by some of the carbon dioxide escaping reduc-
tion, and is mixed with the hydrocarbon gases and vapors
distilled from the coal during its descent to the grate. The
gas passes through the flue A to the main gas flue /.
The gas is enriched by the decomposition of the water,
which is always present, or of the steam blown in, form-
ing carbon monoxide and hydrogen (this mixture is called

water gas).

H^O-^-C^CO + H^

Originally, air was drawn into the producer through the
grate by natural draft, later by steam being blown in with
it. It was soon discovered that a more economical way was
to introduce the air and steam by means of a steam jet, so
arranged that the discharge of the steam draws air into the
producer. A simple form of steam jet commonly used con-
sists of an annular opening that can be enlarged or reduced
by raising or lowering a plunger controlling the opening.
Only a limited amount of steam can be used continuously,
as the reaction forming water gas is so strongly endother-
mic (absorbing heat), that the temperature in the producer
is lowered below the point of reduction of carbon dioxide to
carbon monoxide, the decomposition of the steam thereby
impoverishing the gas by carbon dioxide, and also by steam
passing through the producer to the main. The chief func-
tion of the steam in the ordinary producer (not considering
the manufacture of water gas proper) is to introduce the
air, and at one time blowers were frequently substituted,
but later abandoned for the steam jet. Some recent experi-
ments, however, indicate that the superiority of the latter
over the blower has been much overrated. It is fairly well
established that a large part of the hydrogen in producer

§ 33 MANUFACTURE OF STEEL 35

gas comes from the decomposition of the rich hydrocarbons,
and there is usually enough moisture in the air introduced
to furnish the desired amount of this element.

32. Water-Seal Producers, — The principal improve-
ment in producers since the original Siemens producer was
made has been the adoption of a closed bottom. To accom-
plish this, the producer proper rests in a water pan, through
which the ashes or clinkers are raked out. This water acts
as a seal, preventing the escape of gas and the introduction
of air, which occurred in the old producers while the fires were
being cleaned, contributing much to their irregular working
and the poor quality of gas. Instead of being flat, the grate
is conical, underneath which the pipe conveying the air and
steam terminates, introducing these in the center of the
producer, thus insuring a more even and regular circulation
within the chamber than when they are drawn in at the
side. The air naturally seeks the passage of least resistance
and a serious defect of older producers, where the air and
steam came in at the side, was the tendency to creep up the
walls of the producer without the CO^ first formed being
reduced or the steam decomposed. This also produced
excessive heat, causing the ash to clinker and scaffolds to
form on the side walls. The same conditions may exist
to some extent in any producer improperly managed,
but they are much less liable to occur if reasonable care
is used.

33* Forter Water-Seal Producer. — Fig. 11 shows one
of the most successful and a general type of the water-seal
producer. It is the usual brick-lined shell of steel a. There
are usually but two steam jets s on opposite sides to intro-
duce the air and steam into the wind box w and under the
grate. In this one, a third steam jet s' forces them into the
center of the producer by means of a pipe beneath the ash-
pan, with the vertical part of it terminating below the
grate, as shown at ^, and protected from ashes by a cone-
shaped hood. The wind box has a number of air-tight doors,

3G MANUFACTURE OF STEEL § 33

through which sections of the grate can be removed to bar
out any large cHnkers accumulating on the bottom. The
ashes sHde down into water in thi; ash-pan c as the coal is
burned, and are removed from time to time without inter-
fering with the working of the producer.

34. Frasei*-TallK>t Mechanical Producer. — Recently
a producer has been patented, called, from its inventors,
the Fraser-Talbot mechanical gas producer, in which the
podng or stirring is done by mechanical means. This

g 33 MANUFACTURE OF STEEL 3?

producer, shown in Fig. 12, is essentially the same as
the ordinary water-seal type. A hollow shaft a passes
vertically through
the producer, and
to this radial
arms b are at-
tached. This shaft
has both a rotary
and vertical mo-
tion; the former
revolves the arms
through the mass
of coal, and the
latter constantly
changes the plane
of rotation so that
the horizontal arms
are made to keep
the whole mass
thoroughly broken
up for the passage
of air. The shell
of the producer is
riveted to I beam
columns c, which
extend above the
shell and form a
framework, to
which is attached
the rotating and
lifting mechanism,
which is driven by

an electric motor. '°' "

The central shaft and radial arms are water cooled, as they
are likely to reach a low-red heat and bend from the resist-
ance of the bed of fuel. The advantages are in the quality
and quantity of gas made per unit and the lower cost of
labor. The fire is kept much more uniform than by the

MANUFACTURE OF STEEL

§33

best hand poking, so that the carbon dioxide formed is more
certain to be brought in contact with the carbon and reduced
to carbon monoxide. Holes in which the CO^ can escape
reduction cannot form in the fuel bed from insufficient
poking.

TABIiE U

pm>:^MATE ANALYSIS OF COAI.

Number of
Sample

Volatile
Matter.

Per Cent.

Fixed
Carbon.

Per Cent.

Ash.
Per Cent.

Sulphur.
Per Cent.

36.20

34.70
32.80

33-75

53.20

5»-45
58.10

55.00

5.60

6.85

9. 10

11.25

.85
1 . 00

^ , , . ,

.92
1 .02

•t ••••••••••••••

ULTIMATE ANALYSIS OF COAL

Number of
Sample

Total
Carbon.

Per Cent.

Hydro-
gen.

Per Cent.

Oxygen and
Nitrogen.

Per Cent.

Ash.
Per Cent.

Sulphur.
Per Cent

75-63
76.63

73-92
72.87

4.30

4.57

4-73
4.76

13.62

10.95

"-53
10. 10

5.60

6.85

9. 10

11.25

.35
1 .00

•I

.92
1 .02

"T* * ... ••• •

35. Fuel Employed for Makinsr Producer Gas. — The ,

fuel to make producer gas is bituminous or anthracite coal,
coke, charcoal, peat, or even wood. We will consider only
the first, as the others are of so little importance that they
can be ignored, being used to a small extent only in steel
works and under special or isolated circumstances. The
coal used should be a good quality of gas coal, quite free
from sulphur, having a low or moderate percentage of ash,

§ 33 MANUFACTURE OF STEEL 39

and of such a character as not to clinker on the grate.
While practically all bituminous coals (if not too high in sul-
phur) may be used, there is a decided difference in their
value. Proximate and ultimate analyses of four samples
of good average coal for producer gas are given in Table II.
The former (with the sulphur) is all that is necessary for
the ordinary valuation of a coal for this purpose.

Ordinarily, the higher the coal is in volatile matter, the
richer is the gas produced, as it contains more hydrocar-
bons. Sulphur should not exceed 1 per cent., but this
depends on its condition in the coal — if it is in such a combina-
tion that it is mostly oxidized, remaining.with the ash as sul-
phate, it may be much higher; if principally volatilized, even
this amount may allow the steel to absorb too much of it
from the gas.

36. Producer Beactions. — The reactions taking place
in making producer gas are:

1. Carbon burned to carbon dioxide,

2. Reduction of the CO^ by the hot coal to carbon
monoxide, C(?, + C = 2 C(9

3. Incandescent carbon decomposing water vapor,

11,0 +C=z CO + H,

On the grate in the bottom of the producer are the ashes
which serve to heat the steam and air; and, in connection
with .the water seal, prevent the escape of gas in cleaning
the fires. Next above this is the bed of incandescent fuel,
where the air and steam combine with the carbon in the
above reactions. On top of this is the section where distil-
lation occurs. The temperature is constantly lowered by
the addition of fresh coal, but the heat of the bed beneath
keeps up the distillation of the volatile products of the fuel.
While the ash bed is sharply separated from the one above,
the two upper ones overlap and their reactions occur to a
considerable extent in the same region.

40 MANUFACTURE OF STEEL g 33

The reactions are not all as simple as expressed in the
above equations, as a series of more or less complicated proc-
esses of dissociation and synthesis occur. Under certain
conditions, part of the distillation may take place lower
down in the hotter section, when the original hydrocarbons
will be partly broken up and new ones formed. Accord-
ing to Siemens, some of the carbon deposited in the
regenerators will at that temperature be taken up by the
carbon dioxide and water vapor. This absorbs a large
amount of heat, which is given back on combustion in the
furnace, so that the calorific power of the gas is increased
beyond the increment due to the elevation of the tempera-
ture of the' gas alone. The production of gas is regulated
nearly automatically, as the amount of gas withdrawn deter-
mines the supply of air to the grate — assuming, of course,
that the producer is otherwise properly managed. One
volume of carbon monoxide produced requires 2J^ volumes
of air containing 2 volumes of nitrogen to pass through
the grate, 1 volume of water vapor on decomposition gives
1 volume of hydrogen and 1 volume of carbon monoxide.

31 • Operation of the Producer, — From the preceding
description, the operation of the producer will be readily
understood. The fuel is fed in through a bell and hopper,
by shoveling or by chutes from overhead storage bins. As
the coal becomes hot, it partially disintegrates and cakes,
forming layers, through which the air is forced with difficulty,
or channels are made through the coal so that a large part of
the carbon dioxide first formed will not be brought in con-
tact with carbon and reduced to carbon monoxide. To
avoid this, **poke holes" are placed in the top of the pro-
ducer, through which the incandescent mass is at intervals
of a few minutes broken and stirred with long pokers.
Ashes and clinkers are removed about every other day,
depending on the quality of the fuel and the rate at which
the producer is driven. Other conditions being right, the
hotter and deeper the fire, the better the reactions take
place. The usual depth of fire is about 6 feet, varying with

MANUFACTURE OF STEEL

the ashes on the grate and the rate of feeding the fuel. If
the contents of the fire gets much deeper than this, it is
impossible to keep the bottom of it broken up, however well
it is poked; if much shallower, the carbon dioxide and
water vapor are not decomposed.

38, Composition of Producer Gas. — Under the condi-
tions outlined above, the limits of composition of producer
gas will usually be about as given in Table III.

TABIiE III

Constituents

Minimum.
Per Cent.

Maximum.
Per Cent.

Good Average.
Per Cent.

Carbon dioxide

Oxvfifen

30
.0

18.0

6.0

1.0

58.0

8.0

•5

•5
25.0

12.0

4.0

65.0

5-5
.0

Ethylene

Carbon monoxide . . .
H vdrocren

23.0
8.0

Methane

30
60.5

Nitrocfen

The first two columns are not to be understood as show-
ing analyses of individual samples, but as the usual extremes
of the component gases. Such extreme samples might
rarely be obtained except in the nitrogen, but even this is
exceptional, as the percentage of nitrogen remains quite
constant at 60 to 62 per cent., the variation occurring
mainly with the other gases. Steam is always present in
the gas from some of that introduced with the blast, esca-
ping decomposition, from the moisture, and from the com-
bined water of the coal ; the amount from the first source
depends on the condition of the fire. Tar is always present
in the gas, varying with different coals. It furnishes con-
siderable heat value, which is usually estimated at from
6 to 12 per cent, of the total calorific value of the gas, not
all of which, however, becomes available in the furnace, as

42 MANUFACTURE OP STEEL g 33

part of the tar is precipitated in the gas main, valves, and
flues. The hydrogen comes from the breaking up of the
hydrocarbons and decomposition of the steam. More or less
of the richer hydrocarbons are always decomposed in the
gas tube, producing large quantities of soot, as follows:

This deposition would occur in the hot chambers if not
in the tube ; hence, it is an tma voidable loss, and in the case
of very hot gas fires it becomes excessive. The soot and tar
partially close the gas tube and valves, which must be
cleaned by burning out and scraping at the end of each
week, and frequently require a partial cleaning during the
week.

39. Calorific Value of Producer Gas. — The gas leaves
the producer at a temperature of about 550° C. (1,022° F.)
and is cooled to 100° to 150° C. in the tube. To avoid this
loss of heat, the gas producer has been attached directly to
the furnace, the gas passing from the producer directly to the
ports being hot enough to burn without regeneration. This
seems logical, and is correct from a theoretical standpoint,
but the practical difficulties in the way of its operation have
rendered it ineffectual. From the composition of the gas
given in Table III, the calorific power may be calculated,
but this is of no practical value to the steel metallurgist in
the comparison of different gases, as conditions can seldom
be sufficiently uniform in practice. For practical purposes,
a ton of bituminous coal is taken as yielding 140^000 cubic
feet of gas; this amount, of course, varies with the coal,
the type of producer, and its working. Ordinary producer
gas gives an average of 120 B. T. U. (British thermal units)
per cubic foot, or 1,068 calories per cubic meter. The cal-
culation of the calorific value from the composition does not
show all the heating value in a gas from bituminous coal.
Gas may be made from anthracite coal having the same
composition, but the heating value will be much less, owing
to the absence of solid hydrocarbons in the flame imparting

§ 33 MANUFACTURE OF STEEL 43

luminosity to it. The question of luminosity of the flame
has much to do, in high-temperature work, with the effect
produced. Between a luminous and non-luminous flame in
the furnace, although the actual flame temperature result-
ing from the combustion of the gas may be nearly the same,
there is the difference of rapid melting and entire inability
to reach a steel-melting temperature. This is why anthra-
cite coal will not produce a gas for steel making. At low
temperatures there is little difference between the heating
value of a luminous and non-luminous gas. The incandes-
cent carbon or hydrocarbons cause a large amount of heat
to be given out by radiation. The importance of heating
by radiation in open-hearth steel melting was not recog-
nized for a long time, and the furnace roof was built low, to
confine the flame to the stock. It is now made high, and
the radiative power of the luminous flame is utilized to give
a large amount of the heating effect.

40. Arrangrement of Producers. — Generally the pro-
ducers for an entire plant are connected to one main gas
flue, from which branches, controlled by suitable valves, so
that any one furnace can be cut out without interfering with
the others, go to each furnace. Objections to this arrange-
ment are: (1) The furnaces nearest the producers and
those on the end of the line seldom have the same gas pres-
sure; (2) the deposit, of soot and tar chokes up the tube
nearest the producers, necessitating more frequent cleaning
or a deficient supply ; (3) it is more difficult to maintain a
steady supply than if each furnace has its own producers.

The furnaces at a moderate distance from the producers
receive the best gas ; if too close, the gas is apt to be so hot
that more of the hydrocarbons are decomposed in the regen-
erators, lessening the heating power and increasing the
liability of the regenerators being choked with soot. On
the other hand, if the gas must travel too far, it is cooled so
much that carbon and tar deposit in the cooler part of the
tube, producing practically the same effect as with too hot
a gas.

44 MANUFACTURE OF STEEL § 33

To obviate these and other objections, some recent works
have returned to an earlier plan of making each furnace
independent by building separate producers. A more regu-
lar supply is assured in this way, a furnace not being
affected by the varying demands of its neighbors. The
claim is also made of some economy in labor and fuel, as the
gas supply can be more closely adjusted to the demands of
the melting house.

THE ACID OPEN-HEARTH PROCESS

41. General Remarks. — In the acid process, only stock
containing relatively small amounts of phosphorus and
sulphur can be used, as with an acid slag these impurities
are not eliminated, or at least only to a very small extent.
For this reason, the field of the acid process is limited.

42. Heartli. — The acid- or silicious-lined furnace takes
its name from the silica sand or brick used for making the
bottom or hearth. In almost all cases, a natural sand is
used containing from 95 to 99.5 per cent, of silica, with
2.5 to 3 per cent, of alumina; the remainder consists of
combined water, small amounts of lime, magnesia, and
oxide of iron. All silica sands are not suitable for this pur-
pose, a high degree of purity alone not being sufficient,
much depending on the physical character of the sintered
mass produced. Oxide of iron is the most objectionable
impurity^ as well as the commonest, in sands of the above
percentage of silica.

In ** making bottom," the furnace is gradually heated to
nearly a working temperature, when sand is thrown on the
bottom to a depth of several inches. This is allowed to
sinter when more sand is thrown on in thin layers, sufficient
time being allowed between each addition for perfect setting.
The sides and ends are gradually thickened until the hearth
assumes a saucei;-like shape. The hearth finally has a thick-
ness of from 16 to 24 inches on the bottom and sides; the
latter are carried about a foot above what is to be the level of

§ 33 MANUFACTURE OF STEEL 45

the metal bath. Sometimes two sands of different fusing
points are mixed together, the one so refractory that it will
not soften at the full working temperature of the furnace,
the other softening at a lower heat. By varying the per-
centages, a mixture may be obtained sintering or setting
through a considerable range of temperature. The bottom
becomes so hard that it is not eroded by the stock at the
melting temperature and will resound if struck with a tool.
On this quality largely depends the success of the melting.

43. Cliargre. — The charge will vary considerably at dif-
ferent plants or under varying conditions at the same plant.
It may be all pig iron in the pig-and-ore process, or as low
as 15 per cent, pig iron and the rest scrap. Less pig iron
than this is sometimes melted when coke to furnish carbon
is charged with the stock ; this is exceptional practice, and is
not so sure of producing good steel ; it is therefore resorted
to only where scrap is much more abundant and cheaper
than pig iron. In the pig-and-scrap acid process, the charge
is approximately one-third pig metal and two-thirds scrap.
In general, the charge is so adjusted that when melfed the
bath contains from .3 to .6 per cent, of carbon above the
point designed to tap out on. If too little pig iron is used,
the bath has all the carbon, silicon, and manganese oxidized
before the metal is ready to tap, when it becomes pasty and
oxide of iron is rapidly formed, thus wasting the metal, by
increasing the melting loss. The ferrous oxide forms ferrous
silicate, which scorifies the bottom if the slag is not acid
enough to absorb this additional basic compound. A
further and even more serious injury is the introduction of
oxides into the bath that are difficult to remove and injure
the steel, making it ** wild '* to handle in the furnace and
ladle.

The remedy for too little pig or a heat melting ** low " or
** soft " is simply to add pig iron to the bath — pig up — to give
sufficient carbon and silicon to bring the bath to a boil and
get the necessary temperature to tap the. heat. If too much
pig iron has been charged, no harm is done to the quality of

46 MANUFACTURE OP STEEL § 33

the steel, as there is then a bath high in carbon and possibly
containing some silicon and manganese. These can be boiled
out by the action of the flame alone, or almost universally by
the addition of ore, which hastens the oxidation of the impuri-
ties. The objections to pigging up are (1) time is lost, as
the addition of fresh pig lowers the temperature, the opera-
tion being held back while recovering this heat ; (2) more
pig is required than if the requisite amount had been added
with the initial charge.

In steel works the pig iron is commonly designated as
**hard" and the steel or wrought-iron scrap as "soft"
stock — the terms indicating the relative amounts of carbon.

44. Metliod. of Cbargrinfir* — Generally in an acid furnace
the pig iron is charged on the bottom and the scrap on top.
Sometimes the pig is allowed to heat up, or partially melt,
before the scrap is added. In plants where hand charging
is used, the stock is gradually added, and in the judgment
of many open-hearth managers, the wait between the pig and
the scrap charges gives the men a rest without delaying the
operation. With a charging machine, it is more common to
add all the stock at once — i. e., continuously until all is in.
The usual time of hand charging a furnace of 25 to 50 tons
is from 2 to 4 hours; this may be considered practically a
thing of the past, especially with large furnaces. ' With a
machine, if continuous, from ^ to 1^^ hours is required,
though the time may be extended as long as for hand
charging. The advantages claimed for slow charging are
(I) that the stock has time to heat up as added, and
melting goes on faster ; (2) that the furnace is not chilled
by charging the whole amount of cold stock in a short inter-
val, thereby cooling the waste gases and the regenerators so
that the gas and air are not sufficiently preheated for rapid
melting. Against this view, it is maintained that in slow
charging the furnace doors are up so long a time that a
large amount of heat is lost by the admission of so much
cold air to the melting chamber, and melting is thereby
delayed ; the loss from oxidation is also increased and more

§ 33 MANUFACTURE OF STEEL 47

gas is used. In a properly designed and working furnace,
with ample regenerative capacity, there should be no serious
delay from too rapid charging, there being a sufficient
reserve of heat in the checkers to keep up the temperature.

On an acid bottom the pig metal is charged first, a layer
of it being distributed on the bottom and banks so that the
scrap is kept from contact with the hearth. All of the scrap
is then charged on top of the metal. If the scrap is charged
on the bottom, the waste from the formation of ferrous
silicate is excessive. This basic slag takes up silica from the
hearth until satisfied — i. e., becomes neutral or even acid —
when it ceases to scorify the bottom. This cutting, or scori-
fication, may be a serious matter, as a hole may be started
that will cut entirely through the sand bottom. The sand
will also become impregnated with iron, so that its refrac-
tory power and ability to withstand the action of metal and
slag is lessened. The covering of sand on the pig iron, and
the presence of silicon, carbon, and manganese, by their
oxidation, prevent the pig metal from scorifying the acid
bottom, as would the scrap.

•
45. Calculation of the Cliarfir^* — I- While the calcu-
lation and adjustment of the charge is an important matter,
no fixed rule can be given that can be rigidly adhered to,
as there are so many changing conditions. Chief of these
is the variation in the working of the furnace, causing a
greater or less loss of the elements in melting down. In a
charge for an acid furnace, the composition of the pig is
usually within the following limits :

Silicon 1.25 to 2.00j^

Total carbon 3.00 to 4.00^

Manganese .40 to .80j<

Phosphorus, not over .lOj^

Sulphur, not over .05^

The phosphorus and sulphur depend on the percentage
allowed in the finished steel and the scrap used. Assuming
the phosphorus and sulphur in the stock to be within the

48 MANUFACTURE OF STEEL § 33

limits allowed in the steel, the calculation is based on the
carbon, silicon, and manganese. The value of the latter
elements depends on the oxygen consumed in their oxida-
tion, as shown by the following simple equations:

C+0==CO Si+, (9, = StO, Mn + 0 = MnO

12 -h 16 = 28 28 + 82 = 60 56 + 16 = 71

Expressed in oxygen equivalents for unit parts of the
elements :

(1) 1 part of carbon requires 1.333 parts of oxygen;

(2) 1 part of silicon requires 1.143 parts of oxygen;

(3) 1 part of manganese requires .291 part of oxygen.

Expressed in unit parts of oxygen :

(4) 1 part of oxygen oxidizes .750 part of carbon;

(5) 1 part of oxygen oxidizes .875 part of silicon;

(6) 1 part of oxygen oxidizes 3.438 parts of manganese.

Expressing the other two elements in terms of carbon :

(7) 1 part of silicon is equivalent to .857 part of carbon
(eq. 2 -^ eq. 1).

(8) 1 part of manganese is equivalent to. 218 part of carbon
(eq. 3 -5- eq. 1).

The carbon escapes as a gaseous product, being oxidized
first to carbon monoxide and then to carbon dioxide. The
silicon or silica from the stock forms with the manganese
and iron from the bath a double silicate of iron and man-
ganese, the slag. It may be assumed that in melting down
the stock, from 35 to 45 per cent, of the total carbon in the
charge (silicon and manganese being figured in terms of
carbon) is oxidized. This, of course, can only be approxi-
mated, being affected by furnace conditions, character of
stock, flame, etc. Assuming a loss in melting of 40 per cent,
of the carbon in the charge, the heat to be tapped at .2 per
cent, carbon, it is desired to have it melt at .8 per cent,
carbon, how much pig and scrap of the following analysis
must be charged ?

§33

MANUFACTURE OF STEEL

Elements

Carbon

Silicon

Manganese

Pig Iron.
Per Cent.

3-75
1.50

.60

Steel Scrap.
Per Cent.

.20
.01

•50

Converting to terms of carbon (by equivalents 7 and 8),
we have in the pig iron

1.5 per cent, of silicon X .857 = 1.285 per cent, of carbon

.6 per cent, of manganese X .218 = . 131 per cent, of carbon

The pig contains 3. 750 per cent, of carbon

Total 5. 166 per cent, of carbon

There is in the scrap, disregarding the silicon,

.5 per cent, of manganese X .218 = .109 per cent, of carbon
The scrap contains 200 per cent, of carbon

Total 309 per cent, of carbon

n. The simplest way to treat this matter now is as fol-
lows: As was assumed above, the heat is to melt at .8 per
cent, of carbon with a loss of .4 per cent, of the carbon in

melting ; then .8 per cent, is (100 — 40) or 60 per cent, of

g
the carbon required in the charge, then ^ = 1.333 per cent, of

carbon required in the charge. The question now is how much
pig iron with the equivalent of 5.166 per cent, of carbon and
scrap steel with the equivalent of .309 per cent, of carbon is
required to give a charge with 1.33 per cent, of carbon ?

Subtracting the mean (1.333 per cent.) from the percent-
age of carbon equivalent in the pig iron gives the number of
parts of scrap required.

5.166 - 1.333 = 3.833 parts of scrap.

Subtracting the percentage of carbon equivalent in the

scrap from the mean gives the number of parts of pig iron

required.

1.333 — .309 = 1.024 parts of pig iron.

60 MANUFACTURE OF STEEL § 33

Hence, 3,833 pounds of scrap must be charged with 1,024
pounds of pig.

3,833 + 1,024 = 4,857, or the total charge.

The pig equals j^|^ of the total charge and the scrap
equals i^i^ of the total charge. The calculation may be
completed by proportion or expressed by percentages.

If we have a charge of 75,000 pounds, then by proportion,
letting ;r = weight of pig required,

4,857 : 1,024 = 75,000 : x;

X = 15,800 pounds.

Letting y = weight of scrap required,

4,857 : 3,833 = 75,000 : y\

y = 59,200 pounds.

Solving by percentages, we have

What per cent, of 4,857 is 1,024 ?

(1,024 X 100) -T- 4,857 = 21 percent, of pig.

What per cent, of 4,857 is 3,833 ?

(3,833 X 100) -7- 4,857 = 79 per cent, of scrap.

This latter method is to be preferred, as the charge is usu-
ally figured to a percentage basis.

III. For another charge, where a high-carbon steel is
wanted; the heat to tap at .8 per cent, carbon and to melt
40 points (40 per cent.) above this, or at 1.2 percent, carbon.
Allowing a loss of .35 per cent, in melting down (with the
higher carbon in the charge the percentage of loss will be less,
though the amount of carbon lost may be as high or higher),
we have 1.2 per cent, carbon -^ .65 (1.00 per cent.— .35 per
cent.) = 1.846 per cent, of carbon to be in the charge (silicon
and manganese are figured in equivalent of carbon). How
much of the same metal and scrap used in the previous heat
will be required ? According to the first method, we have

5.166 — 1.846 = 3.320 parts of scrap;
1.846 — .309 = 1.537 parts of pig iron.

§33

MANUFACTURE OP STEEL

Removing decimal points, 3,320 4-1,637 = 4,857 parts
represent the total charge, of which the scrap is 3,320 parts,
or (3,320 X 100) -f- 4,857 = 68 per cent.; the pig is
1,537 parts, or (1,537 X 100) -r- 4,857 = 32 per cent.

rv. For another charge, suppose different stock must be
used.

Elements

Pig Iron

Carbon
Equiva-
lent.

Per Cent.

Rail-Steel
Scrap

Carbon
Equiva-
lent.

Per Cent.

Carbon

Silicon ....
Manganese

3.4oji X I.ooo

3.15^ X .857

.40jiX .218

3.400

2.700

.087

.45Ji X I.ooo
.15^ X .857
.90^^ X .218

•45

•13
.20^

Total . . .

6.187

.78

Assuming a loss of 45 per cent, of the metalloids in melt-
ing, the heat to melt at .9 per cent, carbon, then
.9 -7- .55 (1.00 — .45) = 1.64 percent, of carbon in the charge,

we have

6.187 — 1.64 = 5.097 parts of scrap;

1.64 — .78 = .86 part of pig.
Proceeding as above, we have

(5,097 X 100) -J- 5,957 = 85.5 per cent, of scrap;
(860 X 100) -7- 5,957 = 14. 5 per cent, of pig.

In the preceding calculations the sulphur and phosphorus
were assumed to be such as to produce a steel within
the limits called for. Both are beyond control in the acid
process, the entire amount in the stock going into the finished
steel, and hence are readily calculated from the stock and the
steel specifications. The lower the sulphur and phosphorus
in the stock, the higher is its cost, making it economical to
use the least amount of the purer stock required to finish the
steel within the required specifications. This will always
apply to materials purchased, but in the case of a works

52 MANUFACTURE OF STEEL § 33

using scrap from another department, it will not "generally
be a consideration.

46. Methods of Heating:. — Heating in general is accom-
plished by two methods: (1) By direct contact of fuel and
substance, as a piece of iron in a smith's forge or the coke
and ascending gases in a blast furnace in direct contact with
the rest of the stock ; (2) by radiation, as the heating of a
room by a grate or stove or heating in a mufBe furnace. In
the open-hearth furnace melting is accomplished by both
direct contact and radiation. In the early open-hearth con-
struction it was the practice to build the roof very low, or
even depressed, so as to keep the flame close to the stock
and bath. This was later abandoned and the roof made
higher, allowing free space for combustion.

47. Melting: the Charg^e. — During the time of char-
ging, heating up, and melting the charge, it is usual to
carry a ** smoky *' flame, or a comparatively reducing one,
less air being admitted than is necessary for complete com-
bustion. By this means the charge, especially the scrap, is
kept from oxidizing, the pig being largely protected by its
impurities. This smoky flame is partially self -regulating as,
coming in contact with the cold stock, the temperature is
lowered sufficiently to precipitate out part of the carbon
before combustion takes place. As already stated, the port
construction should be such as to admit the air above the
gas. So far as melting is concerned, this is mainly to keep
next the metal a stratum of gas, instead of air which would
increase the oxidation. This also keeps the flame from the
roof and a relatively cooler stratum next to it. Irregulari-
ties on the slopes of the ports, from neglect on the part of
the furnace helpers in leaving holes or allowing pieces of
brick, etc. to accumulate, may deflect currents of gas or air
either vertically or horizontally, so that the flame is streaked,
and sections of it may be either strongly oxidizing or redu-
cing, or part of the flame may be directed against the roof
or sides of the furnace; even small tongues of flame may

§ 33 MANUFACTURE OF STEEL 63

start cutting of the roof which soon becomes serious if
neglected.

The melting is, in the main, an oxidizing action, though
more or less of the oxide of iron formed may later be
reduced by coming in contact with carbon, or silicon, and
manganese, if the two latter are in the bath. The metal-
loids are removed to some extent simultaneously, but sili-
con and manganese are first oxidized during the melting:
down stage, or immediately thereafter. Generally, only
about one-third of the carbon is oxidized in melting, owing
to its smaller affinity for oxygen under the conditions. In
case a charge was made up of stock very low in silicon and
manganese stock, more of the carbon would be attacked
while melting; or if very high, more of the two former
elements would be left after melting. A certain percentage
of silicon is necessary in the charge that the proper slag
may be formed and to produce heat by its oxidation.

48. The function of the slag is to form a blanket or
covering for the bath, protecting it from oxidation and
transmitting the heat, together with the oxygen, for the
removal of silicon, manganese, and carbon. No definite
rule can be given for the amount of slag that should be
allowed, but it should be thick enough to protect the metal
and not so heavy as to offer too much resistance for the
heat and oxygen to reach the bath. An acid slag will usually
represent from 6 to 10 per cent, of the weight of the charge.
This varies with the percentage of silicon and manganese in
the charge and the conditions of melting and working of
the furnace. The slag is nearly self-adjusting, or is so
within quite narrow limits; that is, a charge too low in sili-
con (or silica) will have this deficiency supplied by the basic
slag formed taking up silica from the hearth. If it contains
too much silica, this will be corrected by the absorption of
iron from the bath. Both are objectionable, as the first
scorifies the hearth and may start a cutting of the bottom
that will result in holes, and even at times in cutting
entirely through. Heats have been lost in this way.

MANUFACTURE OF STEEL

§33

The second correction causes excessive oxidation of the

bath and a consequent high melting loss. Typical acid

slags have the composition sho\irn by the analyses given

in Table IV,

TABIiE IV

Analyses

SiOt
Per Cent.

MnO
Per Cent.

FeO
Per Cent.

MnO + FeO
Per Cent.

49-5
47.6
52.2
46.2

16.5
12. 1

23.4
20.6

30.0

36.3
22.5

28.7

46.5
48.4

45-9

49-3

2 • .. . .

•t* .•.....•.

From the above table it will be noticed that the sum
of MnO and FeO is quite constant. The silica does not
vary over wide limits, and the necessary bases are governed
by the character of the stock. If a charge is low in manga-
nese, the required bases in the slag will be made up by a
larger percentage of ferrous oxide, or vice versa. Analy-
sis 2 shows a slag from a heat with low manganese in the
stock. Analysis 3 is a slag in which a high manganese
stock was melted.

49. Bemoval of tlie Metalloids, Etc. — In Art. 46, the
oxygen-consuming power of the metalloids is given. This also
approximately shows their affinity for oxygen, and the order
in which they are oxidized, which is as follows: First, man-
ganese ; second, silicon ; third, carbon. The manganese and
silicon are first oxidized simultaneously during the melting-
down stage, though traces of both may remain to the last of
the carbon. With an excess of silicon in the charge and the
temperature very high, this order of oxidation may be partly
changed. At very high temperatures, carbon is oxidized in
preference to silicon, the latter remaining in the bath.
This cannot happen in the open-hearth furnace to the extent
possible in the Bessemer converter, as the same high tem-
perature is not reached during the oxidation of the silicon,

§33

MANUFACTURE OF STEEL

as this takes place much slower in the former. If the
amount of silicon and manganese in the charge is more than
is required by the oxygen that can be taken up during
melting, then the excess of both elements remains in the
bath. If ore is added, they will be oxidized before the car-
bon is acted on; but if boiled out by the action of the flame,
the carbon will be removed along with, or partially before,
the silicon. Table V shows the reduction in carbon, silicon,
and manganese in two heats.

TABIiE T

Number

First Heat

Second Heat

Test

Carbon

Silicon

^^"^ Cai
ganese

rbon

Silicon

Man-
ganese

I. GO

1.28

.30 I

1 .60G

.40

I. GO

1. 12

.18 I

.910

.20

I.GG

•51

.09 I

.260

.06

I .OG

'33

.04 I .

. 140

trace

I .OG

•33

trace i

.o8g

I .GO

•05

.G20

•

.90

.02

.GI5

.8g

trace

.28

•55

.IG

.44

,OG

•25

.90

.18

.68

60. Addition of Ore. — When stock that is too high in
carbon is melted, ore is added to hasten the oxidation of
the metalloids. In ordinary practice this means only the
oxidation of carbon, as both silicon and manganese will
have been removed before the bath is ready for oreing.
However, if the latter remain at this stage, they are first
attacked before the carbon is appreciably acted on, if at all.

56 MANUFACTURE OF STEEL § 33

The ore used is a red or specular hematite as free as possi-
ble from all impurities. An analysis of an ore used is as
follows: Iron, 65.6; silica, 2.4; phosphorus, .03; manga-
nese, .3.

Ore may vary somewhat from this analysis and be suit-
able, but as the oxide of iron is the effective agent, the
higher the ore is in this, the greater is the amount of work
that will be accomplished by a given weight of ore. It is
essential that it be in lumps and of sufficient specific gravity
to sink through the slag and the bath of metal, so as to
reach the point where its work — the oxidation of the metal-
loids— is to be done. If in a fine condition or of a low
specific gravity, part or all of it may remain in the slag with
little benefit to the bath, while it will at the same time
increase the amount of slag.

The following reactions take place during ore additions:

Fefi^ -f 3C = 3C(7 + %Fe

Fefi^ + 3i^/;/ = ZMnO +2Fe

Quantitatively, in the relation of oxygen and metalloids,
these reactions correspond to those given in Art. 46.
In addition to its oxidizing action, each molecule of Fefi^
liberates 2 atoms of iron; or 160 parts, by weight, gives
112 parts of iron. This is added to the bath, thereby
increasing the yield of metal by that amount. By some
authorities, it is held that the iron reduced from the ore is
only partially added to the bath, the most of it going to the
slag. This is purely a theoretical point and of little
moment, for, as a matter of fact, if the slag requires oxide
of iron, it will take it either from the bath or as it is released
from the ore, possibly preferring the latter; but if ore is
not added, the necessary oxide of iron will be taken from
the bath, consequently the metallic iron reduced from the
ore may be assumed as a net gain.

61 • From the reactions given, the weight of ore required
to oxidize a given percentage or weight of carbon, man-
ganese, or silicon can be readily calculated:

§33 MANUFACTURE OF vSTEEL 67

1(50 parts, by weight, of Fe^O^ oxidize 30 parts of carbon;
160 parts, by weight, of Fefi^ oxidize 42 parts of silicon;
160 parts, by weight, of Fefi^ oxidize 156 parts of manganese.

In practice this is not done even approximately, as condi-
tions in the melting vary to such an extent that any calcu-
lation is likely to be worse than useless. If the bath is hot,
the ore is acted on rapidly so that the flame has little chance
to contribute its share of the oxygen ; if the bath is cold, the
ore must be added in small quantities, as it lowers the tem-
perature very considerably; under this last condition the
oxygen from the flame will effect the greater part of the
oxidation. Besides, the action of all heats is not the same ;
variations in stock, gas, slag, etc. introduce conditions
that make even approximate calculations of little value.
However, it may be broadly stated that 2,500 pounds of ore
will oxidize the carbon in a 75,000-pound charge from 1 to
.1 per cent.; or 250 pounds of ore will oxidize the carbon
.1 per cent, in such a charge. This is only an approxima-
tion, and about as close a one as can be given. Any silicon
or manganese present has the ** right of way " over the car-
bon and must be first satisfied by the ore. In case of a bath
high in carbon, the ore first added is much less efficient in
oxidizing it than at a later period. This may partly be due
to the last traces of silicon and manganese, and partly to the
condition of the slag, as its viscosity with high carbon
retards the action of the ore.

In Table VI, which is taken from Campbell's ** Open-
Hearth Process,*' in the Transactions of the American
Institute of Mining Engineers, August, 1893, the average
amount of ore used in boiling down a series of heats and
the oxidation of silicon, manganese, and carbon in oreing
are given.

From a study of this table it will be noticed that the amount
of ore is not governed wholly by the percentage of carbon
in the bath after melting. Other conditions that affect it
are the temperature and the way the heat takes the ore,
as the physical conditions of the bath and the slag influence

MANUFACTURE OF STEEL

§33

the reduction effected by a given amount of ore. The judg-
ment of the melter determines when ore should be fed, and
this may not be done at the proper time, so that a series
of tests, however accurate, may be affected by a num-
ber of circumstances other than the quantitative work done
by the ore. In Table VI is shown one heat melting at
.36 p^r cent, carbon, requiring no ore to bring it to .08 per

TABIiB VI

Elements or Metalloids

Pounds of Ore Used

in the Heat

I 090

.54
.08

.02

.09
.04

850

.64
.08

.05
.01

.06

.02

None

.36
.08

.03
.02
.06
.04

500

.18
.08
.01
.01
.03
.02

t,ooo

.32
.08

.04
.03
.05

.02

x.Soo

.61
.08
.07

.02

.15
.05

3,000

Per Cent, j After melting. . .'

Carbon i Before tapping

Per Cent j After melting

Silicon ( Before tapping

Percent ( After melting

Manganese ( Before tapping

.57
.08

.09

.02

.15
•03

cent, carbon, and another heat melting at .18 per cent, car-
bon requiring 500 pounds of ore to bring it to .08 per cent,
carbon. This is explained in the one case by the tempera-
ture being too low to work the ore, the flame affecting the
oxidation ; and in the other by the bath being so hot that
the ore is rapidly reduced. The last two heats show con-
siderable silicon and manganese when melted, which will
account for part of the ore.

53. nnishlng: the Heat. — In Table VI the analyses
show the steel to contain .08 per cent, of carbon in all cases
before tapping. If soft steel is wanted, it is necessary to
boil down to this point, or nearly so. In the harder grades
of steel (those higher in carbon), if other conditions are
right, the bath may have the carbon but slightly reduced
below the amount desired in the steel. In making the
soft and medium grades of steel, those below .4 per cent,
carbon, for example, the bath is either boiled down to about
that shown in Table VI, or is stopped when just below the

§ 33 MANUFACTURE OF STEEL 59

steel specification, or, as it is called, caught coming down. In
the first case, any additional carbon that may be required
is furnished by the recarburizer or recarbonizer. There are
certain advantages in both methods, and the subject will be
treated under the heading **Recarbonization." At whatever
percentage of carbon the heat is to be tapped, it is essential
that the temperature be right at the same time. The tap-
ping point might be represented by a given point, and the
temperature and carbon content as lines^ or forces, approach-
ing it from different directions, the object being to have the
two strike this point at the same time. The melter controls
both within very close limits by an adjustment of the flame
and the feeding of the ore. With a hot bath and relatively
high carbon, ore would be fed rapidly; with the same per-
centage of carbon and a relatively cold bath, ore would be
fed slowly or not at all, depending on the conditions.

63. The most essential requirement in a skilful melter is
his ability to read temperatures accurately. No apparatus
is used for determining this, the eye alone^ with the aid of
ordinary blue glasses to cut off the intense heat and light
rays, shows it within very close limits. The relative and
not the actual temperature is determined, as for all practical
purposes this answers fully as well. It is necessary to esti-
mate the temperature of both the melting chamber and the
bath. The former is shown by the flame, slag, and, mainly,
by the appearance of the side walls and roof. The tem-
perature of the metal can be ascertained only by reaching it
direct, and other indications are frequently misleading. The
more common method is to try the heat by inserting an iron
rod into the bath and stirring it back and forth, noting the
rate at which the rod melts; or stir it for a given time,
usually ^ or 1 minute, withdrawing it, and observing the
way the metal has cut the rod: a clean, sharp end melted to
a point indicates a hot bath, while a colder bath will melt
the rod much less, but more regularly, rounding it off, for
the rod will be built up by the mushy, thick metal. The rod
must be thrust quickly through the slag, or the latter will

60 MANUFACTURE OF STEEL § 33

coat and protect it from the action of the bath, so that the
indications given by the test will be misleading. In the
hands of an experienced melter, the " feel " of the metal as
the rod is stirred back and forth gives an idea of the tem-
perature, as it is more limpid and of less viscosity when
hot. The surface of the bath will sometimes be as hot as
desired, while portions of the bottom will be pasty from
partially melted stock.

Another way is to take out a sample of the metal in a
small test ladle and pour it into 2t mold or into a cake on the
floor. The character and temperature is shown by the way
it pours; its fluidity, or viscosity; the sparks given off; the
skull remaining in the ladle ; the contraction of the test on
cooling; and general indications that are easily learned in
practice, but which cannot readily be described. This test
piece is also used to determine the amount of carbon, either
by fracture or from drillings taken from it for a rapid-color
carbon test (see Quantitative Analysis),

If the tests are carefully taken and uniform conditions
observed in cooling, an experienced eye can usually read the
carbon, as shown by the fracture, within 2 or 3 hundredths
of a per cent, in samples under .2 per cent, carbon. Above
this, as the carbon increases, the error in judging by frac-
ture also increases. These tests are taken at intervals until
the proper percentages of carbon and temperature are
reached, when the tapping hole is opened and the metal run
into the ladle. The proper recarbonizers having been added
in the furnace or in the ladle, the metal is poured into molds
in a pit or on cars.

THE BASIC OPEN-HEARTH PROCESS

54. Introductory. — The basic process, either the open-
hearth or the Bessemer, differs from the acid process in that
stock higher in phosphorus and sulphur is treated and basic
materials, usually lime, are added, to give a slag that
will effect purification. As previously explained, the only

§ 33 MANUFACTURE OF STEEL 61

difference in the apparatus used is that the hearth is
made of a basic instead of a silicious material. The idea
should be clearly grasped that the hearth performs no
office in effecting the purification — the dephosphorization
and desulphurization — the basic slag alone being account-
able for this work. It is necessary to have the hearth
either of a 'basic or neutral material, so that the slag will
not react with it.

56« Advautaifes of the Basic Process. — The advan-
tages of the basic process are that a wider range of stock is
made available for steel making, that purer steel may be pro-
duced, and cheaper stock used. These two statements might
seem to be conflicting, as a better or purer material would
not be expected from inferior stock. This view retarded the
growth of the basic process to a great extent, as many users
of steel refused to believe that steel made from impure
materials was as good as that made with purer stock. This
view, however, is now held by scarcely any one either among
the producers or users. Rarely does an engineer specify
acid steel to the exclusion of basic for important uses ; one
or two notable exceptions have recently come up where acid
steel only was allowed in important engineering work, and
this must be taken as the judgment of an individual engineer
rather than the accepted or proved practice. The only
objection that can now be raised to basic steel is the impure
stock used, but the process effecting purification does not
leave this a valid one. This was not always so, as defects
in the process and manipulation caused the steel to be defect-
ive, and much of the earlier prejudice against basic steel
was founded on fact. The present methods of manufacture,
however, both from a metallurgical and engineering stand-
point, make basic steel equally as well adapted as acid steel
for practically every purpose. The furnace, except the
hearth, and all accessories are identical with the acid
process, and the steel is made from pig and ore or pig
and scrap, with a lime addition, with or without ore, as
in the acid.

62 MANUFACTURE OP STEEL § 33

HEARTH MATERLAXS

56. Xeutral Materials. — It is immaterial whether the
hearth is of neutral or basic material, but in present practice
it is altogether the latter, and this is all that need be con-
sidered. The neutral materials that have been used in
hearths are carbon in bricks or mixed with refractory
materials ; bauxite ; and chromite. None has been entirely
successful.

Carbon is unsuitable mainly because of the affinity of the
metal for it. It is readily absorbed — the hearth thus being
gradually destroyed. It would be an ideal material to resist
the action of the slag, but the above objection renders its use
out of the question.

Bauxite is one of the most refractory substances known,
but its excessive shrinkage at high temperatures causes it to
crack and thus unfits it for this purpose. It is practically
neutral under all conditions. It thus has two most essential
points. It has been thoroughly burned and shrunk before
being used, but this, by causing loss of combined water,
destroyed its plasticity, which is important.

Chromite is highly infusible and withstands basic condi-
tions in a high degree. In fact, the chief point against it is
its inf usibility, as it is difficult to sinter or set a bottom with
it, so that erosion takes place, owing not to lack of refrac-
toriness, but to the mechanical condition in which a hearth
is left.

57. Basic Materials. — The strictly basic materials for
the hearth are lime, dolomite, and magnesite.

Lime is the cheapest and most widely distributed material;
it occurs in the form of limestone, or calcium carbonate
CaCO^, Theoretically, burned lime, or calcium oxide CaO,
is well suited for hearths, but practically it does not answer,
as it slakes so rapidly on exposure to the air that it cannot
be kept in stock. A bottom made of it when heated would
partially crumble into dust, owing to the driving out of the
water and gas, and would be rapidly worn away by the
metal.

§ 33 MANUFACTURE OF STEEL 63

Dolomite^ or magnesian limestone CaMg{CO^^y was orig-
inally much used owing to the high price of magnesite. It
is abundant in many and relatively cheap in all localities,
and when thoroughly burned does not absorb enough mois-
ture to slake for some time. It has been used with tar,
rosin, or other material to bind it until set by the heat. The
tar is generally discarded now and the material thrown in
without any binding agent. It has been made into bricks
and the bottom built up with them. Bottoms have also
been made by ramming in loose layers. The best method,
however, is the same as making up a sand bottom, by
sintering • in thin layers, allowing time for each stratum to
be thoroughly set.

Magnesite^ or magnesium carbonate MgCO^^ when cal-
cined to MgOy is the ideal material for basic hearths so far
as our present knowledge of refractories goes. Practically
all hearths now put in are made of it, although many dolo-
mite hearths are still in use. Its high cost barred and
retarded its use for a number of years in the basic process,
but discoveries of large deposits in Austria and Greece have
lessened the cost greatly. The Grecian magnesite is much
the purer, and is generally considered to make the better
brick, but it is not adapted for making bottoms, as it is too
refractory when used alone. To lower its fusing* point by
the addition of silica, clay, or oxide of iron is too uncertain
in results and does not give a bottom having as good phys-
ical qualities to resist wear and erosion as the calcined natural
Austrian magnesite. Bottoms are wholly made of the latter
and the patching done with it. The bottom is made the
same as one of dolomite or silica, by setting successive layers
and generally using a little basic slag to make it flux ; clay
may be used in place of slag, but the latter is preferable.

On the bottom of the basic hearth generally two courses
of magnesite brick are laid or one of magnesite and one of
chromite brick. This is done to offer greater resistance to
the metal or slag should the bottom be cut through. The
side walls also are built of magnesite brick until near the top
of the lining, sometimes only to the foreplate, or two or

64 . MANUFACTURE OF STEEL § 33

three courses above. Silica brick are used above the mag-
nesite in the side walls and for the roof. Formerly it was
considered necessary to have a neutral or passive joint
between the two, as it was held that the silica and magnesite
would flux. Any of the neutral or passive substances above
mentioned answer, but chromite is best adapted. The idea
that they will flux in the side walls has been proved errone-
ous, and silica brick are laid directly on the magnesite brick
with no neutral body between. It is only essential that the
silica walls be protected from the basic slag, and this is
provided for by the bottom of magnesite being carried
on the sides and ends above the slag level when the charge
is melted.

68. Charge. — In regard to the metal, the charge differs
from an acid charge only in that more pig iron can be,
and usually is, melted. This is owing to the fact that the
carbon dioxide CO^ from the limestone acts as an oxidizing
agent on the elements in the bath and also that there is less
objection to mixing ore with the original charge, so that
more oxidation is effected during the melting-down stage.
Limestone, or burned lime, is added with the charge to form
the basic slag. Technically, there is no difference which is
used, so far as forming a basic slag and removing phosphorus
is concerned, but the furnace is the cheapest place to burn
the stone; hence, the raw limestone is almost universally
used. The pig iron should be as low in silicon as possible, a
maximum of 1 per cent, is the highest allowed in good practice
and usually it does not exceed .75 per cent. As each pound
of silicon in the pig iron requires, roughly, 15 pounds of
limestone, the importance of having the silicon at the lowest
possible point is apparent. The above ratio is only an
approximation, as silica may come from other sources, and
the percentages of phosphorus and sulphur largely deter-
mine the amount of lime to be charged. A large lime charge
is objectionable from its increased cost ; but especially as it
means an increased amount of slag, so that the time of melt-
ing is lengthened, cutting down the output of the furnace;

§ 33 MANUFACTURE OF STEEL 65

extra fuel is used to form the slag and afterwards to get the
heat through the heavy covering; it is harder on the fur-
nace, as the fine dust is carried against the silica roof and
over into the checkers — cutting the one and clogging the
other.

69, There is somewhat greater variation in the method
of charging than in the acid process. In the best practice
all the limestone is charged on the bottom, the pig iron is
placed on this, and then the scrap. Some prefer to charge
part of the scrap on the bottom, then all or a part of the
limestone, the pig iron, and the remainder of the scrap last.
Others charge only a part of the stone, and as slag begins
to form from the oxidation of silicon and manganese, add
burned lime as needed to keep the slag sufficiently basic.
The chief advantage with the lime on the bottom is the
better protection it affords the latter ; also, as the stone is
decomposed, the CO^ and CaO coming through the pasty
mass help mechanically to bring action to the bath. The
only objection to placing all the stone on the bottom is that
it sometimes sticks to the basic lining, partially filling up
the melting space. With proper attention from the furnace
men, there should be no serious trouble from this source.
A rod is used to loosen the lime as it begins to ** come off
the bottom."

In recent practice, molten pig metal taken directly from
the blast furnace or from a ** mixer " has been used with entire
success and the practice is being adopted wherever blast
furnaces are operated in connection with basic open-hearth
furnaces. The use of hot inetal^ as it is called, is not adapted
to the acid open-hearth, as the silica hearth of the latter is
rapidly scorified by charging either the molten iron, or steel
scrap, directly on the bottom. In the basic process the
bottom is protected by the limestone and then by whatever
steel scrap is used. The molten pig iron is poured in from a
ladle, carried by an overhead traveling crane, on top of the
rest of the charge. The advantage of hot metal is that
heats are made in much less time, as the melting time is

66 MANUFACTURE OF STEEL § 33

gfeatly lessened, thus increasing the output per furnace.
The scrap is usually heated until it begins to **drip,"
or the metal may be poured in soon after the scrap is
charged.

60, Calculation of tlie Charge. — The weights of pig,
scrap, stone, and ore vary with local conditions, the char-
acter of the stock, and of the steel to be made. Whether
pig or steel scrap is the more abundant or cheaper deter-
mines the percentages of these within quite wide limits
-T-from a minimum of 30 to a maximum of 70 per cent, of
the one, or the other may be used in ordinary practice.
The more pig used, other conditions being the same, the
more limestone is required to keep the slag basic from the
silicon to be oxidized; or the higher in silicon, the more
stone. Phosphorus and sulphur also require lime for their
absorption ; the purity of the limestone largely determines
the amount needed. If high-carbon steel is wanted, more
carbon must be charged, which in this case is pig iron.
The ore is determined by the metalloids to be oxidized ; a
high pig charge means increased ore, and a minimum of pig,
no ore.

Besides the above relations being considered independ-
ently, allowance must be made for their relation to each
other ; i. e. , a large amount of stone and ore cannot be charged
together, owing to the excessive foaming produced. The
percentage of manganese present influences the amount
of CaO required.

The following is a charge for a basic open-hearth furnace
of 90,000 pounds capacity:

45 per cent, of pig iron will equal 40,500 pounds.

Analysis op the Pig Iron

Silicon 76^

Sulphur 055if

Carbon 4.00^

Phosphorus eOjif

Manganese 75^

§ 3^ MANUFACTURE OP STEEL 67

55 per cent, of steel scrap will equal 49,500 pounds.

Analysis op the Scrap

Silicon trace

Sulphur 06^

Carbon 12^

Phosphorus 10^

Manganese 50^

8 per cent, of limestone will equal 7,200 pounds.

Analysis op Limestone

Silica 1.000^

Calcium carbonate 95.7005if

(Calcium oxide) (53.600)j^

Ferric oxide and alumina .800^ .

Magnesium carbonate "ZAOO^

(Magnesium oxide) (1.150)j^

Phosphorus 006j^

Sulphur trace

2 per cent, of iron ore will equal 1,800 pounds.

Analysis op the Iron Ore

Silica 2.500^

Iron 67.500^

Alumina 950^

Phosphorus 042j^

Sulphur trace

Calcium and magnesium oxides 300^

The total charge usually includes only the pig iron and
scrap, but sometimes the iron content of the ore used is
figured in. A portion of the pig iron is usually replaced
with cast-iron scrap, owing to the lower cost of the latter.

Owing to the great variability of conditions, no exact
rule can be given for calculating the charge. It seldom
happens that all the stock is sufficiently uniform to get
more than an average analysis of it. This is generally the
case with scrap, but also to some extent with the pig
iron, limestone, etc. In the charge just given, in order to

r>ft MANUFACTURE OF STEEL § 33

show the calculation, it is assumed that the materials are
uniform.

In the slags given in Table VIII, the proportion of cal-
cium and magnesium oxides to silica is very variable. Such
wide divergences are due to the other elements in the slag
and to the conditions of melting. But, fortunately, even
with the rest of the composition the same, the ratio of cal-
cium and magnesium oxides to silica may vary greatly, so
that no exact calculation is necessary, or even possible.
The basis of the calculation is the silica, calcium oxide,
and phosphorus. The phosphorus becomes calcium phos-
phate Ca^{PO^^ and ferrous phosphate Fc^{PO^)^ in the slag,
but sufficient calcium oxide is allowed for all the phos-
phorus. Assuming that this is done, we have ^CaO to 2/^, or
168 parts, by weight, of calcium oxide to 62 parts, by weight,
of phosphorus; or 2.7 pounds of calcium oxide to 1 pound
of phosphorus, this being merely the theoretical amount
required for the reaction. In practice, about 3 pounds of
calcium oxide is allowed for 1 pound of phosphorus. Some-
what more calcium oxide is allowed for the silica, about
4 pounds to 1 pound of silica. Applying this to the actual
working charge just given, we have the following cal-
culations:

Calculation for Phosphorus

40,500 lb. pig iron at .6^ phosphorus = 243.0 lb. phosphorus
49,500 lb. scrap at .1^ phosphorus = 49.5 lb. phosphorus

Total charge contains 292.5 lb. phosphorus

(The small amount of phosphorus in the ore would be
disregarded.)

292.5 X 3 (the ratio of CaO to P) = 877.5 pounds of cal-
cium oxide required for the phosphorus.

Calculation for Silicon

40,500 pounds of pig iron at .75 per cent, of silicon
= 303.75 pounds of silicon. S/ : 5/d^, = 28 : 60, or the
weight of silicon x 2| = weight of silica.

§ 33 MANUFACTURE OF STEEL 69

303.75 lb. silicon in pig iron X 2| = 650.9 lb. silica
7,200.00 lb. limestone at 1^ SiO^ = 72.0 lb. silica
1,800.00 lb. ore at 2.5^ SiO^ = 45.0 lb. silica

Total charge contains 767.9 lb. silica

767.9 X 4 (the ratio of CaO to 5/(7,) = 3,071.6 pounds of
calcium oxide required for the silica.

Calcium oxide required for the phosphorus = 877. 5 lb.
Calcium oxide required for the silica = 3,071.6 lb.

Total calcium oxide required for silica

and phosphorus = 3,949.1 lb.

To find the amount of limestone required, MgO is fig-
ured as CaO\ therefore, the stone is considered as contain-
ing 53.6 + 1.15 = 54.75 per cent, of available CaO, Then,
3,949.1 pounds -7- .5475 = 7,213 pounds of limestone required;
or, in practice, 7,200 or 7,225 pounds would be taken.

In the first experiments, trouble was encountered in keep-
ing up the bottom, but the preceding method of charging
was adopted and little or no difficulty results. From two
to four heats extra per week can be made by using hot
metal, which results in an increase of from 15 to 25 per
cent, in the output.

61, lilme Addition. — The function of the lime, as
already explained, is to form the basic slag by which the
dephosphorization and desulphurization are effected. The
amount of lime required depends primarily on the amount
of silicon or silica in the charge ; and after satisfying the
SiO^ with an excess of lime, a further basicity is required to
remove phosphorus and sulphur, depending on the percent-
ages of the latter elements present. From 90 to 98 per
cent, of the phosphorus in the charge is removed. Sulphur
is more difficult and uncertain to control; frequently over
half is readily removed, while, again, when conditions
appear almost the same, a reduction of 10 per cent, will be
hard to obtain.

70 MANUFACTURE OF STEEL § 33

It might seem that a basic slag is all that is required, and
if made so from iron, this should effect purification. This
would be objectionable from an economic point, but tech-
nically because a slag high in ferrous silicate, i. e., rich in
FeO^ has its iron readily reduced when in contact with a
bath high in carbon, so that a slag sufficiently basic to keep
from scorifying the bottom could not be maintained. This
principle of the ready reducibility of ferruginous slags is
availed of in the Talbot and Monnell open-hearth processes,
the former especially making a beautiful application of this
reaction. It is therefore necessary to have bases that will
not be reduced, as lime or magnesia. The latter has been
used, but is not so effective, as a slag high in magnesia is
less fusible, more viscid, and refractory (which means more
fuel), and is harder on the furnace. Lime, either as lime-
stone CaCO^ or as burned lime CaO^ is the essential basic
addition. Economy determines in which form this lime-
stone is added, but it is almost always used as the raw
stone. The use of the latter affects the process by the
carbon dioxide liberated by the decomposition of the car-
bonate. This carbon dioxide acts as an oxidizing agent on
the metalloids of the bath, thus allowing a larger percentage
of pig iron to be used, which is an advantage when this is
cheaper stock than steel scrap. The following reactions
show the relation of the carbon dioxide as an oxidizer:

C-\-CO^-%CO

Mn + CO^ = C0 + MnO
Fe+CO, = CO + FeO

By some, the carbon monoxide working through the
partly melted mass is said to cause foaming^ a frothy action
of the slag due to gases passing through it. Foaming is not
only caused by carbon monoxide, but also by other gases,
and by silicon under certain conditions of temperature and
working. There is then danger that the metal and slag
may be carried over into the ports, boil out the doors, or
that the slag may come in contact with the silica side walls

§ 33 MANUFACTURE OF STEEL 71

of the furnace, cutting these and introducing silica into the
slag. About the only remedy for foaming is by checking
the action of the bath — if from carbon, shutting off the gas
until the action lessens; if from silicon, making the slag
more basic by the introduction of burned lime.

6JJ, Use of Ore. — Ore is used in the basic just as in the
acid process, both by charging with the stock and by feed-
ing after melting, to oxidize the carbon, etc. The amount
charged depends on the percentage and character of the pig
iron used and how low the carbon is to be boiled down.
The reactions of the ore are as follows :

3C + Fe^0, = SCO + 2/>
SSi + 2Fe^0, = dStO, + 4/>
SMn + Fefi^ = ZMnO + 2/>

The reactions do not take place immediately, as there are
a number of intermediate steps, but the ultimate results are
the same. As has been stated, the carbon monoxide causes
foaming and limits the amount of limestone and ore that can
be charged. Both the carbon dioxide from the stone and
the ore Fefi^ are reduced by the carbon of the bath, the
other metalloids being first oxidized (Art. 60). The ore,
however, produces less carbon monoxide than does the lime-
stone, for the same amount of carbon oxidized.

(1) Limestone,

CaO'CO^ + C = 2C(9 + CaO

100 +12= 56 + 66

(2) Ore,

/>,(?, + 3C = SCO + %Fe
160 + 86 = 84 +112

From the above equations it is seen that for every atom
of carbon oxidized by the limestone, or more strictly by the
carbon dioxide, 2 molecules of carbon monoxide are pro-
duced ; while in the case of ore 3 atoms of carbon produce
only 3 molecules of carbon monoxide. Or in the first case,
each carbon atom gives 2 volumes of carbon monoxide ; in

72 MANUFACTURE OF STEEL § 33

the second, each carbon atom shows only 1 volume of carbon
monoxide produced. So that for a given amount of carbon
monoxide produced, twice as much carbon is taken from the
bath with ore as with stone. Nearly as much difference is
shown in their oxidizing effects, as 100 parts of limestone
take 12 of carbon, while 160 parts of ore take out 36 parts
of carbon [see equations (1) and (2) above]; 53J parts, by
weight, of ore accomplishes the work of 100 parts of lime-
stone, or the ore is 1 J times as efficient an oxidizer of carbon.

63, There is a great difference, also, in thermal condi-
tions, resulting from the reactions shown by the above
equations. The first, reducing carbon dioxide to carbon
monoxide, is endothermic (absorbing heat) ; the second is in
two phases, endo- and exothermic (liberating heat) ; the
first phase consists in reducing the F^^0^\ and in the second
phase heat is developed when the oXygen reduced from the
ore combines with the 3 atoms of carbon. The second
phase produces more heat than the first phase absorbs,
so that the net result is a gain in heajt.

Limestone has been termed a refrigerating agent ^ owing
both to the distillation of its carbon dioxide and the action
of this on the metalloids to form carbon monoxide. The
terms refrigerating agent and calorific agent as applied,
respectively, to limestone and ore must not be taken too
literally, for in practice these effects do not stand out so
prominently as the above might indicate. That the facts
are as stated can be proved by calculations of the heat
absorbed and developed by the reactions given. In practice,
this may be modified or obscured by other factors, but the
net results are as given.

64, Melting:, Etc. — Melting on the basic hearth is an
oxidizing action in the main with the same forces at work as
on the acid. In addition, there are several relations changed
or modified, and the essential difference of a basic slag car-
ried, to effect the removal of phosphorus and sulphur.
Art 46, I, gives the oxygen-consuming power of the

§ 33 MANUFACTURE OF STEEL 73

metalloids. In basic practice, phosphorus is added to
the list according to the reaction

62 + 80 = 142

One part, by weight, of phosphorus unites with 1.290 parts
of oxygen; or 1 part of oxygen with .775 part of phosphorus.
The oxygen-absorbing power is only slightly less than that
of carbon (1.333), or 1 part phosphorus is equivalent to
.9G8 part of carbon, Art. 46, I. This relation of phos-
phorus also accounts for the larger percentage of pig iron
that can be melted in basic practice.

In general, the most easily oxidized elements are first
burned. In acid practice it was shown that the formation
of oxide of iron was necessary to combine with the silica to
form the slag. In basic practice we have the lime to com-
bine with the oxidized silicon, and silica originally in the
stock, so that there is not the same call for iron to be
oxidized, but iron oxide is always present in basic slag.
Just why the necessary conditions cannot be fulfilled by
the other bases is not so apparent. In general, slags seek
to absorb or combine with whatever increases fluidity and
fusibility, and this may explain why ferrous oxide is taken
up, its presence giving greater fusibility. With an increase
of lime in the slag, the percentage of iron decreases, as a
rule, but there are a number of conditions modifying this.
The amount of manganous oxide MnO and phosphorus
pentoxide P^O^ greatly affect the fluidity of the slag and
lessen the necessity for ferrous oxide.

65. The matter of viscosity of the slag is of the utmost
importance in basic open-hearth work, and is a function of
the composition and temperature. A too viscid slag will
not readily transmit the heat and oxygen of the gases to
the bath, so that the oxidation of the metalloids is delayed,
while a too fluid slag will cut the basic hearth, even if the
excessive fluidity is not due to silica, although the latter is
usually the cause. The remedy for such a slag is to render

U MANUFACTURE OP STEEL 1 33

it basic, and this must be done promptly, for at the high
temperature it rapidly attacks the hearth. Burned lime or
dolomite may be used, but the former is much better and
is almost always employed, as magnesia renders the slag
viscid and requires greater heat for the same fluidity.
Frequently the lime comes up and remains on the surface
without dissolving. This condition, of course, is due to a
deficiency of silica, for the lime to readily combine with; it is
generally an advantage rather than otherwise — i. e., within
the limits of sufficient silica in the charge to form a slag with
the bases. To cut up the lime in such a case, or to render
a too basic slag more fusible, fluorspar, calcium fluoride CaF^^
is employed by throwing a few shovelsful (from 25 to
200 pounds) on the slag or lime.

Silica or a silicate will, of course, thin the slag very
quickly, but it is a remedy that may cut both ways and
attack the lining or lower the basicity of the slag, so that
dephosphorization will not take place completely or allow
some of the phosphorus to return to the bath. Fluorspar is
much more efficient, and gives fluidity without lessening the
basicity of the slag. The reaction is rather obscure, but
the most probable explanation is the formation of a double
fluosilicate. Manganese ore is sometimes used for the same
purpose and is very efficient ; it has the additional advan-
tage th^t the manganese oxide in the slag acts as a desul-
phurizing agent also. The increased fluidity from man-
ganese is not due to any reaction with the silica or lime,
except that a more fusible compound is introduced into the
slag. This latter may also partially explain the action of
calcium fluoride.

66. Basic Open-Hearth Slagr. — Chemically, the slag is a
silicate of calcium, iron, and manganese. Magnesia is always
present both from the limestone, and the dolomite or mag-
nesite of the hearth and that used for patching; the amount
furnished from the limestone is almost always much less than
that from the other sources. Alumina is also present in
amounts usually varying from 2 to 6 per cent., its source

§33

MANUFACTURE OF STEEL

is the same as magnesia ; these two compounds are not to be
considered essential, but rather incidentally present from the
nature of the case.

No fixed limits can be given for the composition; as
previously stated, the two essentials are fluidity and basicity.
The former first, that it may flow freely from the furnace
with, or immediately after, the metal, so as not to fill up the
hearth; second, that the reactions may take place without
too much resistance from the slag, so that the **boir' will
not be checked when the metalloids are oxidizing. Basic-
ity is necessary, first, to remove the phosphorus and sul-
phur of the charge ; second, to preserve the basic lining of
the hearth.

The ordinary limits of composition of a good slag are given
in Table VII.

TABIiE

VII

Limit

5/(7,

CaO + MgO

FeO

MnO

PtO,

Minimum. ..
Maximum . .

lO
20

lO
25

5
15

If the silica runs much below 10 per cent, the slag is too
viscid to properly perform its function, unless sufficient
fluidity is furnished by liquefying elements, especially man-
ganese and phosphorus. If above 20 per cent., there is
always danger of cutting the bottom and a failure to purify
the bath, but in case of very low phosphorus in the charge
the silica may exceed the maximum given without harm.
High ferrous oxide generally goes with low silica, and vice
versa, but there are many exceptions to this. The calcium
and magnesium oxides will depend on the other bases and
on the phosphorus pentoxide, though primarily on the per-
centage of silica. The MnO and phosphorus pentoxide
result from the manganese and phosphorus in the charge.
By a consideration of the conditions existing, it will be seen
that ferrous oxide is the only compound that the slag has

MANUFACTURE OF STEEL

§33

TABIiE Vm

Number of
Slag

5/0,

CaO

FeO

MnO

AO.

MgO

AUO^

18.30

50-25

14.91

4.85

3-43

6.00

1.96

14.60

50.04

10.20

7.15

6.50

8.07

2.62

19.20

45.62

9-04

9.60

2.98

10.28

3.28

2943

3907

14.61

8.00

5.28

5.20

8.70

51-90

23-45

7.25

11.00

3-70

12.20

41.20

18.30

5-30

12.60

6.40

14.25

39-97

13.18

10.84

9-51

8.49

9.85

43- 46

14.81

8.26

15.38

4.23

8.50

45-30

18.25

8.00

12.40

4.50

10.90

42.70

12.09

10.26

13-70

5-58

The above are all representative slags except No. 4, which
is exceptional from the high percentage of silica. It is given
because such slags are occasionally met with, but imperfect
purification of the metal or excessive scorification of the
hearth generally results. The sulphur exists in the slag as
sulphide, principally as CaS^ the percentage usually being
from :i^ to 1 per cent. CaS,

69. In conclusion, it may be said that the essentials of
the slag are silica, calcium oxide, and ferrous oxide in pro-
portions to give a fluid, basic slag. The presence of man-
ganous oxide is highly desirable to give fusibility and to
desulphurize, but it is not an essential; it is, however, always
a constituent, depending on its percentage in the stock.
The phosphorus pentoxide is present as a result of the
oxidizing action and the basicity of the slag; this is the
fundamental principle of the basic process.

Even under the best conditions of working, the slag always
scorifies the hearth somewhat. The slag line, or shelf ^
requires patching after each heat, burned dolomite or mag-
nesite being used, generally the former, as it is cheaper,
also as it sets more quickly and is thus more permanent.

§ 33 MANUFACTURE OP STEEL 79

Holes frequently are left in the bottom after the heat is
tapped; the metal and slag must be bailed or splashed out of
these with rabbles (heavy iron hoes) and the holes filled with
magnesite, a little slag usually being added to increase the
fusibility, so that it will set quicker.

The amount of slag produced in basic work is necessarily
much greater than in acid, and usually ranges from 8 to 20 per
cent, of the weight of the charge. This depends on the
amount of slag-forming elements in the stock, the amount
and quality of limestone used, and the melting practice.

70. I>epliospliorlzation. — The removal of phosphorus
takes place partially during the oxidation of the other metal-
loids, as a rule, or may be complete before the carbon is all
burned, but the greater part of the manganese and all of the
silicon is oxidized before dephosphorization can be finished.
It is probable, under certain conditions, that the phosphorus
is simultaneously oxidized with the silicon. Owing to the
conditions of melting, it is practically impossible to obtain
data that will accurately set limits within which dephosphori-
zation occurs. The* essential thing, of course, is to have
sufficient lime present to form not only a basic slag, but to
leave enough in excess to absorb the phosphorus pentoxide
formed. The phosphorus in the slag exists as a phosphate
of iron or calcium. The purpose and exact office of ferrous
oxide, with respect to phosphorus, is not understood, but a
certain amount seems to be required.

71. In melting stock low in phosphorus, the elimination
may be essentially complete during the melting period, while
with high phosphorus the percentage of removal will be
much less, though the actual amount may be greater. There
is no relation between the amount eliminated during melt-
ing and that present in the charge ; with apparently uniform
conditions as to stock, conditions of melting, etc., wide
variations are shown in practice. Some of the practical
obstacles in determining the dephosphorizing conditions
referred to above are the uncertainty as to whether all
lime has come off the bottom when the heat is melted ; the

MANUFACTURE OF STEEL

§33

kind and arrangement of the stock; the character of the
flame; changes in the slag due to irregular stock and vary-
ing percentages of ferrous oxide. Table IX shows the
phosphorus removed during melting, together with partial
analyses of the accompanying slags. These are given as
examples met with in practice and not intended that any
general deductions should be drawn as to dephosphorization,

TABIiE rX

ANALYSES SHOWrN^G ELIMINATION OF PHOSPHORUS

DURING MELTING

a
S

Initial Phos-
phorus in Charge

Per Cent, of

Pljosphorus

After Melting

Amount of

Phosphorus

Removed

Per Cent, of Phos-
phorus Elimi-
nated in Melting

Accompanying Slags

•44

SiOt

FeO

P%0^

CaO

3.00

.755

2.245

75.00

10.41

-i^l'Zl

48.56

2.18

.629

1. 551

71.00

12.79

4.41

23.50

43.07

2.29

.849

I.44I

63.00

12.68

4.05

21.83

42.18

1.42

•563

.857

60.00

II. 10

16.05

•55

.282

.268

49.00

30.26

10.08

5-99

45.26

.55

.297

.253

46.00

31.30

10.98

3.72

4145

.55

.378

.172

31.00

34.05

18.45

3.08

3509

•55

.464

.086

16.00

3437

6.57

.19

.009

.181

9500

13.02

24.21

.19

.032

.158

83.00

14.09

27.09

.19

.072

.118

62.00

22.93

II. 16

.19

.105

.085

4500

25.34

6.66

as many conditions affecting this cannot be shown. The
first four given show unusually high phosphorus in the
charge ; in this case the elimination was increased by a large
amount of ore charged with the heat. The second four
show a much lower elimination with a smaller percent-
age of phosphorus in the charge ; in this case a deficiency of
lime, as shown by the high silica in the slags, will mainly
account for this.

§ 33 MANUFACTURE OF STEEL 81

The last four show a low phosphorus charge for basic
practice with elimination nearly complete in one case ; these
also show the removal greater with lower silica in the slag.

72. When the heat is melted, a test is taken and broken
to show the melter where the carbon is and if the phosphorus
is low. The latter can be told by the fracture, the same as
the carbon, but with much less certainty, so that this is gen-
erally determined by analysis. After the lime is all up, the
melter adjusts the slag (entirely by the eye); if too acid,
burned lime is added to bring it to the proper consistency ;
if too basic, it is thinned with fluorspar or manganese
ore. With the proper slag, if much phosphorus remains in
the bath after melting and the carbon is not too low, it will
generally be oxidized by the time the carbon is boiled down
to the desired point. In case the phosphorus should not be
removed when the carbon is practically all out, it is usually
necessary to add pig iron, which lowers the temperature and
brings action on the bath by introducing metalloids to be
oxidized. This procedure is mainly required, however,
because the slag covering a carbonless bath rapidly takes
up ferrous oxide, but the presence of carbon neutralizes this
action or reduces the Fe from any ferrous oxide formed.

73. The thermal conditions accompanying oxidation of
phosphorus favor its removal during melting, as it enters
the slag at a comparatively low temperature. This does not
mean it is not removed at a high heat also.

In good basic practice, the phosphorus is reduced to less
than .04 per cent, in the finished steel, and not infrequently
in the regular practice the steel shows but .01 to .02 per
cent, of phosphorus. Depending on whether high or low
phosphorus stock is melted, this shows an elimination of
90 to 99 per cent. Table X shows the percentage of phos-
phorus in six samples of finished steel, that in the charge
and in the slags, and represents current practice.

74. Desulphurlzation, — Throughout the manufacture
of iron and steel, sulphur is the most difficult element with

MANUFACTURE OF STEEL

§33

which the metallurgist has to contend. One by one the
others have been controlled and a way found for their elim-
ination, generally by surrounding them with such conditions
that they are made to do useful work. The reactions
involved in their removal furnish a large amount of heat in
all the processes, and in some, all of the heat used in con-
verting the liquid pig iron to steel. In the acid Bessemer
process the oxidation of the silicon, carbon, and manganese
gives all the heat required ; in fact, it may furnish too much
for proper working. In the basic Bessemer, the oxidation

TABIiE X

ANALYSES SHOWING PHOSPHORITS IN FINISHED STEEL.,

ELIMINATION, ETC.

J2
g

Initial Phos-
phonis in Charge.

Per Cent.

Phosphortas in
Ingot.

Per Cent.

Phosphorus
Eliminated.

Per Cent.

Accompanying Slags

o
d

Per Cent,
of FeO

•

Per Cent.

3.000

.036

98.8

13.98

4.68

16.62

52.73

I 350

.040

97.0

8.28

1347

"•37

6.98

55.77

.190

.012

94.0

1504

21.02

.100

.005

95.0

18.30

15-30

3.43

4.85

56.25

I.OOO

.015

98.5

12.20,

4.80

.820

.020

97.6

10.60

of phosphorus is the chief source of heat. In both the acid
and the basic open-hearth processes the oxidation of the
impurities furnishes a large amount of the heat. Sulphur is
the only one of the ordinary impurities in pig iron that has
not been fully utilized in some process. Both basic methods
remove a part of it, but cannot be said to control it, as
results are irregular and any large reduction cannot be
counted on with certainty.

75. Manganese and lime are the only agents, chemically,
that are used ; temperature is a potent factor in eliminating

§ 33 MANUFACTURE OP STEEL 83

it, and there seems to be no limit, except what the furnace
will stand, at which a high heat is not an advantage.
The action of manganese and lime is as follows:

(a) Manganese effects reduction (1) by that which is pres-
ent in the stock, carrying out sulphur with it as oxidized ;
(2) by the use of manganese ore, which, as reduced, adds
manganese to the bath, acting as ia (1), or, if incorporated
directly in the slag, is reduced from the latter during decar-
bonization and dephosphorization ; (3) by the addition of f er-
romanganese or spiegeleisen to the bath, which act as in (1),
but are more effective for the same amount of manganese.

(d) A limy slag absorbs sulphur, the only conditions
being extreme basicity and high temperature. In connection
with lime, calcium chloride has been used ; its use is covered
by a patent, the process being called the Saniter process^ from
its developer. At the time of its introduction (1892) great
claims were made for its efficiency and much was expected
from it. It did not reach any general application, and is
not used today in America, and by only a few works in
England, where it originated. In using calcium chloride, an
unusually limy slag is carried, the function of the chloride
apparently being to furnish fluidity, the extra basicity of
the slag likely taking care of the sulphur without any
direct help from the calcium chloride. Any other agent
that increases the fusibility without lowering basicity, thus
allowing a more limy slag to be carried, is as efficient.

Fluorspar assists in desulphurizing in the same way. It is
improbable that it has any direct action, but by giving the
necessary physical condition to an otherwise too viscid slag
it may be classed as an indirect desulphurizer.

76. With high sulphur in the charge more is removed,
under otherwise similar conditions, than with low sulphur
stock. This is apparently due to the greater tenacity with
which smaller percentages of all the elements remain in the
bath. . The non-uniformity of removal is mentioned above.
As a general statement, it may be said that one-third of the
sulphur in the charge is eliminated in good basic practice.

84 MANUFACTURE OF STEEL § 33

This is without any special effort or losing time for adjust-
ing the slag by any of the additions given above. By the
latter course, a removal of from 50 to 75 per cent, can be
effected regularly; and this is frequently reached in reg-
ular working, without particular plains, except as noted,
but cannot be relied on. Whether it is an economy to get
a high elimination of sulphur depends on the cost of the
purer stock, as the more sulphurous, the more time is con-
sumed, thereby reducing the output. The extra basic addi-
tions are harder on the furnace, on account of dust carried
over to the ports and checkers and the higher working
temperature generally employed. The cost of the extra
additions is of some moment, but usually less than the two
preceding points.

77. Manganese effects renwval of sulphur by metallic
manganese, whether added as such or reduced from ore by
the action of silicon and carbon, taking the sulphur from its
combination with iron, or solution in the bath, forming sul-
phide of manganese, this mostly being absorbed by the slag.
It may also occur that part of the manganese sulphide is
decomposed by the oxidizing action of the slag and expo-
sure to the flame, the sulphur burning to SO^ and the man-
ganese returning to the bath to again take up sulphur, or it
may form manganous oxide at once. Lime may combine
with sulphur directly in the presence of carbon or by react-
ing with the manganese sulphide.

The reactions of lime and manganese with sulphur are as
follows :

1. Mn + FeS = MnS + Fe (absorption by metallic

manganese).

2. J/;/S+ O^ = SO^ + Mn (loss of sulphur in waste

gases).

3. 5 + CaO + C = CaS -\- CO (direct combination of sul-

phur with CaO),

4. MnS + CaO = CaS + MnO,

That part of the sulphur is oxidized and lost in the waste
gases (it seems most probable as shown by reaction 2) is

§33

MANUFACTURE OF STEEL

indicated by the fact that the sulphur irf the slag and the
finished steel does not always account for all in the initial
charge. If using producer gas, from .005 to .015 per cent,
of sulphur will be absorbed .from this source. Natural gas
does not increase the sulphur.

Considerable sulphur is generally lost during melting, but
no regularity attends this. In endeavoring to get accurate
data as to sulphur, more obstacles are in the way than with
any other element. Some of these are : The sulphur absorbed
from the gas; that lost by volatilization; the difficulty of
obtaining the exact amount in the charge, as it will vary more
than any other element, and unless elaborate sampling of the
stock is done, there is greater discrepancy ; most important
and exerting the greatest influence, are changes and influ-
ences that are not fully understood, and from the nature of
the case seem impossible to control. Among the latter are
variations in the slag, character and arrangement of the
stock, temperature, some of the melting conditions, etc.

TABLE XI

Calculated
Sulphur in Charge.

Per Cent.

Sulphur After
Melting.

Per Cent.

Sulphur in
Ingot.

Per Cent.

Per Cent.
Eliminated

.085

.070

.050

41 .2

. 120

. 100

.045

62.5

.070
.280

.050
.220

.020
.086

71.4
693

.060

.040

.030

50.0

.050

.030

.025

50.0

.040

.030

.025

37-5

.045

•035

.030

33-3

•035

.030

14.3

Table XI gives results from regular basic practice of sul-
phur elimination without any attempt to analyze the various
causes in individual cases.

MANUFACTURE OF STEEL

(PART 2)

THE BESSEMER PROCESS

1. Introductory. — The Bessemer process for the man-
ufacture of steel was invented by Henry Bessemer, and pat-
ented in England in 1855. In recognition of his services to
metallurgy and for the far-reaching effects of his invention,
he was afterwards knighted and is generally spoken of as
Sir Henry Bessemer. It is doubtful if any single invention
or discovery has had such a wonderful effect on industry
and manufacturing in general. While it became the basis of
the modern steel industry, in itself of great magnitude, it is
in the development of other industries, made possible by the
cheapening of steel, that we see its full importance. The
railroads in particular, in their present development, could
become a reality only when it was possible to produce the
steel necessary for the rails and other parts of the equipment.
Steamships and engineering and manufacturing establish-
ments of all kinds are made of it or depend on it for success.
In fact, our whole industrial and commercial life may be
said to be more dependent on steel than on anything else ;
but until the invention of the Bessemer process it was impos-
sible to produce steel in sufficient quantities or at a suit-
able cost to permit its general use. While the Bessemer
process has a great future, especially in the United States,
where vast quantities of suitable ores are available, it is
being superseded to a great extent by the basic open-hearth
process; and whenever the cost of production by the latter

§34

For notice of copyright, see page immediately following the tiUe page.

d MANUFACTURE OF STEEL § 34

becomes equal to or lower than the Bessemer it will sup-
plant it still further.

Bessemer experimented several years before taking out
his first patents, which covered the principle of blowing air
through or over molten iron. Many other metallurgists had
worked on the line of introducing a blast of air to effect the
refining of pig iron, but only one other, so far as known, used
a vessel or converter and blew the air from the bottom
through the liquid iron. While the priority of Bessemer *s
invention has been questioned, there is no doubt that his
work was prosecuted independently and that he was the
first to completely realize the full success of the principle.
The other inventor who used this principle was William
Kelly, an American, who experimented about the same time
as Bessemer, and applied for a patent in 1857 ; while Besse-
mer had secured American patents nearly a year before this,
the patent office allowed Kelly's claim on the ground of
priority of discovery. For several years following this, two
companies, representing the Bessemer English and the
Kelly American patents, attempted to introduce the process
into the United States. Litigation resulted and a compro-
mise was finally effected by the former company taking
70 per cent, and the latter 30 per cent, of the United
States royalties. This was only partially in recognition of
the Kelly patent, as the latter company had acquired the
United States rights to the patents of the Mushet recarbon-
izing process by the use of spiegeleisen or ferromanga-
nese. This is the only recognition either Kelly or Mushet
received for their work on the pneumatic process of making
steel, although, unfortunately, the financial rewards of the
above compromise did not reach either. Bessemer's appa-
ratus, from a mechanical point of view, was much superior
to Kelly's, and it was largely owing to this fact that it super-
seded it.

Bessemer's experiments covered almost every conceivable
method of applying the pneumatic principle — blowing from
the top and sides on to the metal or through it ; in various
kinds and types of fixed and movable vessels, etc. He

§ 34 MANUFACTURE OF STEEL 3

finally adopted the tipping vessel or converter with the air
blown through the metal from the bottom. This type of
apparatus, as Bessemer developed it, remains the standard
today. Many mechanical improvements tending to increase
the speed and convenience of working have, of course, been
made, a large number being developed by Alexander L.
Holley, when the process was first applied.

Bessemer's original idea was to produce wrought iron,
but owing to the large amount of gases left in the metal and
the lack of a fibrous structure, the blown metal was worth-
less, as in iron. At this point his experiments rested for
some time and seemed almost a failure. Mushet had previ-
ously added and patented the use of manganese in the form
of dioxide, under reducing conditions, and later as an iron
and manganese alloy called spiegel, or spiegeleisen, for the
production of steel. It was when Bessemer availed himself
of this method that steel was first made by the Bessemer
process.

The first plant built on a commercial scale was at Shef-
field, England, the home of the crucible-steel industry.
AVhile the chief technical difficulties had been so far over-
come as to produce merchantable steel, there still remained
the commercial ones of introducing it and overcoming the
prejudice of users to a new metal, and setbacks from failures
from putting it to uses for which it was not intended, as
sometimes, to replace soft iron or harder steel. Its first use
was in certain tools, machine parts, etc., and later, in ship
building, railroad construction, and the varied kinds of mer-
chant-steel shapes, bars, etc. The great consumption of
Bessemer product has been and yet is steel rails. As the
process grew in England, it extended to the Continent and
the United States.

3. Apparatus Used. — The essential appliance and the
one representing the Bessemer or pneumatic principle, is the
converter or vessel in which the molten pig iron is trans-
formed from cast iron into steel, or more correctly into
blown metal — the recarbonizing being necessary to give the

MANUFACTURE OF STEEL

§34

Fig. 1

final product, steel. As
necessary adjuncts, are
cupolas for remelting,
or some means for ta-
king the pig iron direct
from the blast furnace ;
and the necessary
cranes^ ladles^ molds^
etc. for handling the
iron used and the steel
produced. The move-
ments of the cranes,
converter, etc. are con-
trolled by hydraulic
power, an accumulator
keeping up the pres-
sure, which is usually
maintained at from
500 to 700 pounds per
square inch.

3. Cupolas. — The

cupola furnace used is
shown in Fig. 1. It is
the same as the ordi-
nary foundry cupola
except that it is larger
and is usually placed
at a higher level, owing
to a different method
of handling the iron.
The number and size
vary with the iron to
be melted, but the
usual Bessemer plant
has from three to six.
The height from the
bottom to the top of

§ 34 MANUFACTURE OP STEEL S

the stack is approximately from 40 to 60 feet. The diam-
eter of the shell (of about |-inch riveted steel plate) is
from 10 to 15 feet and is lined with firebrick, giving a
melting space of from 8 to 12 feet in diameter in
freshly lined cupolas; they may rest on solid foundations
or on iron pillars e. All cupolas are arranged with drop
bottoms d^ and are, therefore, always set up some dis-
tance above their foundations to allow the doors of the
drop bottom to swing down. These are necessary to facili-
tate the cleaning out of the cupolas at the end of each week,
or of tener ; they run without repairs from 3 to 6 days, when
they become partially filled up and scaffold across, so that
the space in the shaft is too small for proper melting. About
25 feet above the bottom of the cupola is the charging floor,
where the stock is elevated to be charged into the charging
door/. As this metal melts, it drops to the bottom, whence
it is tapped at intervals into the ladle. As in all melting
furnaces, slag forms; this is removed through a slag hole
at a higher level than the iron notch, or tap hole ; the space,
or well, between the two holes allows the accumulation of
melted iron, so that quite large weights are available at one
time.

4. Tuyeres. — These are the openings (shown at a) in
the shell and lining, through which air is supplied for the
combustion of the fuel effecting the melting. They are either
separately connected to the blast pipe or open from a com-
mon wind box c extending around the cupola. They are
about 4 inches in diameter on their outer ends and taper to
about 2 inches on the inside. They may vary from this size,
but if less than 2 inches in diameter, they are too easily
clogged up by slag and iron. Their number varies also,
but an ordinary 10-foot cupola (inside) will usually have
from 12 to 20 but sometimes as many as 48 tuyeres. The
blast is furnished by a blower, or fan, usually placed in the
engine room, and has a pressure of 8 to 14 ounces per square
inch, as shown on a gauge b\ the tuyeres or main blast pipe
are provided with slides or valves for regulating the volume

MANUFACTURE OF STEEL

§34

and pressure of blast to suit the various melting conditions,
such as the working of the cupola or the amount of iron
wanted. The amount of air used is a variable quantity,
but is approximately 30,000 cubic feet per ton of iron melted.

5. liining:. — The cupola lining consists of the best grade
of firebrick and varies from a thickness of 18 to 24 inches
at the bottom, where the greatest wear and pressure come,
tt) 12 inches in the upper part. The abrasive action of the
descending charge of pig iron, etc. requires a brick of
special quality. The brick are laid up with a thin grout of
ganister'and clay; patching is done with ball stuff of the
same. The lining is built up straight or drawn in towards
the top, as this construction lessens the disposition to hang
or scaffold.

6. Fuel. — This consists almost universally of coke, but
anthracite coal is used as a partial substitute where it is
cheaper; coke, however, owing to its more open structure,
which permits the ready passage of the blast and keeps the
shaft open, is much superior. The coke should be hard-
burned, with strong, firm structure to bear up the burden and
not crush. In composition, sulphur is the most injuri-
ous element and should be as low as possible ; ash is objec-
tionable merely as an adulterant, lowering the melting value
and requiring more flux; phosphorus is of less consequence
than in blast-furnace practice, as the ratio of coke to metal is
so much less and it is doubtful if much or any of the phos-
phorus in the coke enters the iron in melting. The range
in analysis of a good cupola coke is given in Table I.

TABIiE I

Minimum. ..
Maximum. .

Ash.
Per Cent.

Fixed
Carbon.

Per Cent.

Volatile
Matter.

Per Cent.

Sulphur.
Per Cent.

8. GO
I2.00

87.00
91.00

.50

I. GO

.75

I.GG

Phos-
phorus.

Per Cent.

.005
•o*5

§34

MANUFACTURE OF STEEL

In starting the cupola, coke for a bed is charged on the
bottom, with enough wood to readily light it, a little dis-
tance above the wind box, or the lower tuyeres; when this
is thoroughly ignited, with the blast turned on, the regular
charging of pig iron and the necessary coke follows. The
fuel and iron ratios range from 1 pound of coke to 8 pounds
of iron up to 1 pound of coke to 16 pounds of iron, good
practice being about 12 pounds of iron melted with 1 pound
of coke.

7. riux. — The flux used to form a slag with the sand
on the pigs, the coke ash, and wear of lining, is limestone.
The slag contains varying amounts of iron, usually from
10 to 12 per cent., and is mainly a silicate of iron and cal-
cium. The usual range of composition of stone and slag is
given in Table II.

TABIiE II

Silica.
Per Cent.

Lime.
Per Cent.

Oxides of
Iron and
Alumina.

Per Cent.

MgO.

Per
Cent.

Iron.

Per
Cent.

Limestone..
Cupola slag.

I to 5
45 to 55

48 to 53
15 to 20

.5 to 2.5
15.0 to 20.0

I to 5
I to 4

8 to 15

8. Cupola Mix. — This is the proportion in which the
different irons are charged to give the required composi-
tion, together with the fuel and flux. The silicon and
sulphur are the only two elements usually figured on, as all
the iron for the Bessemer process is very close to the same
percentage in phosphorus and does not vary widely in
manganese. The usual range of composition of Bessemer
pig iron is given in Table III.

Carbon is not considered in the calculation. With the
phosphorus below the Bessemer limit of .1 per cent., the
sulphur not exceeding .05 per cent., and manganese about
.6 per cent., silicon is the chief element controlling the

MANUFACTURE OF STEEL

§34

mix, though the sulphur and phosphorus are not less impor-
tant ; in fact, even more so, as the process has no control
over these, while the silicon may vary within considerable
limits without serious disadvantage to the product. The
silicon is the principal fuel in the process and is necessary
for this reason. Formerly 2 to 2.5 per cent, was considered
necessary in the pig iron for successful working, but in
present practice about half of this is used. This is due to
more rapid working and to using less steel scrap in the
cupola and vessel.

TABIiE III

Silicon.
Per Cent.

Sulphur.
Per Cent

Phos-
phorus.

Per Cent.

•

Man-
ganese.

Per Cent

Carbon.
Per Cent.

Minimum . .
Maximum. .

•75

2.00

.02
.06

.08
.10

.4
.8

3-75
4.25

In melting, there is a gain of sulphur, by absorption
from that in the coke, so that the metal tapped out contains
from .01 to .03 per cent, more sulphur than did the initial
pig iron; the increase depends mainly on the amount in the
coke, but somewhat on melting conditions and on the lime
charged. There is a loss of silicon in the cupola, as some is
oxidized and enters the slag, the amount depending some-
what on the initial silicon in the pig and the melting con-
ditions— the blast and rapidity of melting. Under similar
conditions, the loss of silicon is greater the higher the
silicon is in the pig; it is usually taken as .2 to .3 per cent.
No fixed rule is adhered to in calculating the charge, as it
seldom happens that different grades of iron are available
in amounts necessary for an accurate calculation. The con-
dition is more common that certain amounts of one or two
irons must be used and the mix adjusted with other irons,
so that the calculation becomes an approximation and can
be made essentially accurate. Assume a loss of .25 per cent.

§ 34 MANUFACTURE OP STEEL 9

of silicon in melting and that the metal should go to the
vessels to be blown, with 1.25 per cent, of silicon, or the
cupola charge averages 1.5 per cent, of silicon ; that the stock
available requires the use of 75 per cent, of a 1.2-per-cent.
silicon iron or .9 per cent, of silicon from this source; that
the highest silicon iron at the metallurgist's disposal is
2 per cent., so that using the remaining 25 per cent, of this
iron gives .5 per cent, of silicon, or 1.4 per cent, in the mix. A
plus or minus error of .1 per cent, from the desired amount
is permitted. The stock in buggies is raised to the charging
floor by an elevator or lift, hydraulic power being generally
used, and dumped in the charging doors, the coke dis-
tributed between metal charges, and the stone thrown on
the coke barrows at the scales below ; from 40 to 60 pounds
of limestone is added per ton of pig iron.

9, Mixer. — Fig. 2 shows a section through the mixer.
This is a reservoir for storing the molten metal from blast
furnaces, and has come into general use only within the
past 3 or 4 years, but has been used at a few works
for a number of years. The construction is simple, it being
merely a strongly framed structure of steel plates lined'
with firebrick. Two or four hydraulic cylinders c are placed
at each corner, or one side, for tipping it to pour out the
metal. It is provided with a hopper, or funnel a^ at the
back, as shown, or in the center of the roof. The pig iron
is run into ladles at the blast furnace and transferred by a
locomotive to the mixer; or if the blast furnaces and steel
plant are close together, a traveling crane is generally pro-
vided for transporting the ladle and pouring into the mixer.
In the former arrangement the ladle is run up an elevated
track, raised by a hydraulic lift or a crane to a sufficient
height above the mixer to pour in readily.

The advantages in using the mixer are that remelting
the pig in cupolas is avoided, thus saving the expense of
fuel and handling, and that the loss is less. Molten metal
has been taken direct from the blast furnaces to the con-
verters, but the results have not been satisfactory, owing

10 MANUFACTURE OF STEEL g 34

mainly to the frequently varying composition from one cast
to another and also to the fact that the metal was not avail-
able just as wanted, or came in too large quantities when

entire casts came at once. The mixer furnishes the metal
exactly as it is wanted; and what is even more important,

§34 MANUFACTURE OF STEEL 11

it supplies a more uniform metal from the mixing of a num-
ber of casts. These mixers are made to hold from 150 to
250 tons and are of service where blast-furnace and Bes-
semer plants are operated together and there is a large out-
put from both ; generally the metal is transported only short
distances, but it has been successfully taken in ladles from
2 to 3 miles.

Cupolas are usually operated in connection with the
mixers to supply part of the metal for blowing when the
blast-furnace output is below the converting capacity. In
transferring from the mixer to the converter, a ladle on a
car is run under the pouring spout, the mixer tipped over
by the hydraulic cylinders, and the required weight poured
out, as shown by track scales on which the ladle car rests;

MANUFACTURE OF STEEL

§34

the mixer is righted and the ladle moved by electric or
other haulage sys-
tem, so that the
iron crane can pick
it up and pour into
the converters.

10. Converter.

This is the essential
apparatus of the
process and the one
in which the pneu-
matic principle is
applied. It is an
oval vessel with a
symmetrical nose,
as shown in section
in Pig. 3, or an ec-
centric nose, as in
Fig. 4. The former
is more generally
used now, although
at one time the
latter was used
almost exclusively.
It is ma'de of heavy
riveted plate steel
and is lined with
refractory material
— ganister for the
acid, and dolomite
or magnesite for
the basic process, as
in the open-hearth.
It is suspended
about the middle on
trunnions, shown
at //and d'. Fig. 3,

§ 34 MANUFACTURE OF STEEL 13

one of which, d\ is hollow, and through which the blast passes
from the blast pipe by way of the gooseneck e to the vessel's
bottom and thence through the tuyeres to the metal. The
vessel is rotated by hydraulic power applied through a rack
and pinion. The construction is such that it can be made
to revolve completely and empty out any slag after pouring
the steel. Referring to Fig. 3, it will be seen that the ves-
sel consists of three principal sections keyed together to
form the complete converter. The middle, or main section b^
around which the trunnion ring a extends, holds the body
of metal while it is being blown. The bottom ;;/ is detach-
able and is held to the body of the vessel by keys and links /.
Originally, the bottom was not movable, but the latter con-
struction (an invention of Holley's) did much to facilitate
repairs and speed of working. Beneath the bottom proper
is the tuyere box f\ its cover is keyed on at k and is air-
tight ; the nose of the converter c is also keyed on to the main
part, permitting its removal for repairs, etc. The straight-
or concentric-nosed vessels are generally held to slop less
than the eccentric-nosed ones; i. e., less metal is thrown out
of the converters by the violence of the reaction. They are
made in sizes of from 1 to 20 tons capacity, but blow about
6 tons in small plants and from 10 to 20 tons in the large
plants; less than 5 tons is usually for steel-casting plants
where the output is limited. The metal fills only a small
part of the space, as the reaction is so violent that abundant
room must be allowed for it. When the vessel is turned
down, the metal lies in the belly, shown at «, Fig. 4, so as to
be clear of the tuyeres and not run out the nose.

THE ACID BESSEMER PROCESS

11. Introductory* — The acid and basic Bessemer proc-
esses bear the same relation to each other as to the acid
and basic open-hearth processes. The lining for the con-
verter in the acid process being of acid material, dephos-
phorization and desulphurization do not take place, owing to

U MANUFACTURE OF STEEL § 34

the acid slag necessary; hence the process is limited to com-
paratively fine pig irons, as in the acid open-hearth process.

12. Bottom and Tuyei-es of Converter. — The bottom
for the acid process is made up of ganister rammed in or
pieces of the ganister rock set over it and ball stuff of the
same and some clay rammed in between these. Its thick-
ness is 26 to 30 inches. The tuyeres j are spaced over the
bottom and supported from below by the tuyere plate //,
Figs. 3 and 4; they are placed in position before the bottom
is built up and the ganister built up around them. Their
length corresponds to the thickness of the bottom (26 to
30 inches), so that their inner face comes flush with the
latter. They are cylindrical in shape, about 6 inches in
diameter, and contain from 6 to 10 holes f to ^ inch in diam-
eter; their number varies from 7 to 12 and the total tuyere
area (i. e., area of the holes) varies from 2^^ to 4 square
inches per ton of metal blown. After being made up, the
bottoms are run into drying ovens and thoroughly burned.
Their life varies from a single heat occasionally, to 50 or 60
rarely; 30 to 35 heats for a single bottom may be taken as
good average practice. The tuyeres are made of hard-burned
and very refractory fireclay ; in blowing, it frequently hap-
pens that a tuyere will be cut through by the metal — when
the vessel is turned down, the lid ^of the tuyere box removed,
and a circular plate inserted over the tuyere or blanjced; the
heat can be blown with a number of the tuyeres blanked, but
the blowing time is increased. Bottoms are changed in
some works by turning the vessel into a vertical position
with the nose down, and after unkeying, a crane lifts it off
and places a fresh one in position. In others, the vessel is
turned with the bottom down and a car is run under it, on
which the bottom is dropped, a hydraulic lift raising the car
against the bottom ; a fresh bottom on another car is raised
against the vessel and keyed on. The latter is the more
rapid method, but little time is lost by either.

13« lilnlng and Kepairingr. — The lining is about
12 inches thick, and is made up of ganister or silica brick.

g 34 MANUFACTURE OF STEEL 15

usually the former, ground with about a fifth part of refrac-
tory clay. The vessel's lining lasts much longer than the
bottom, as the latter is supporting the charge most of the
time and the cutting action is more intense on it. From
3 to 5 months is an average life for lining, or 6,000 to

10,000 heats; sometimes it may be cut through after a few
heats. Repairs are required constantly, especially around
the nose, which is injured by pouring the steel. Repairs
are made with the regular lining, usually of ganister. After
a vessel has been lined up or patched for the beginning uf

MANUFACTURE OP STEEL

§34

a week, it is thoroughly dried out and made hot before
metal can be poured in.

14. Blast. — This is furnished by vertical or horizontal
blowing engines, generally the former, as they are more
compact. Fig. 5 shows a common type of blowing engine.
The blast is carried in an 8- or 10-inch pipe to the vessel;
control of it is effected by suitable valves controlled by the
blower from the pulpit. A pressure of 20 to 30 pounds f)er
square inch is maintained in the blast pipe, as shown by a
gauge on the pulpit. The pressure is varied according to
the metal to be blown and the conditions of the vessel —
depending on the bottom, number of tuyeres blanked, etc.
The blow lasts from 7 to 12 minutes, but with very large
heats or a deficient blowing capacity it may exceed the latter.

TABIiE IV

T^t A MMM «k ^4

Initial
Pig Iron

Time After Beginning to Blow

Slement

3 Minutes

3 Min. 3oSec.

6 Min. 3 Sec.

8 Min. 8 Sec.

9 Min. 10 Sec

Carbon

Silicon

Manganese..
Phosphorus.
Sulphur

2.gS%
'94%

•43%
.lo%

.06%

2.940%
.630jt
.ogo%

.060%

2.710%

,040%
.Io6jt
.o6ojif

1.720%
-030%
.030%
.Io6jt
.060^

'S30%
.030%

.OlOjt

.060^

.040%
.0205(
.010%

.ioS%

.060%

Character of

Flame From

Converter

Silicon
Flame

Brighten-
ing.
(Carbon
Starting)

Moderate

Carbon

Flame

. Full

Carbon

Flame

Flame
Drops

16. Chemical Changres In the Converter. — In general,
the elements are oxidized in the same order as in the open-
hearth process. In the acid Bessemer process silicon is first
burned to 5/0„ then manganese to MnO^ and, simultane-
ously with this, some iron is oxidized, forming the slag with
the SiO^ and MnO, The silicon and manganese go largely
together, the silicon first under ordinary conditions. The
carbon is next oxidized with ordinary pig iron ; the silicon

§ 34 MANUFACTURE OF STEEL 17

and manganese will be reduced to little more than traces
before much carbon is burned, but with excessively high
manganese the carbon will be largely burned before the
manganese is gone.

Table IV shows the progressive removal of the elements
in blowing.

A study of the table shows that the carbon burns very
little until the silicon and manganese are practically gone.
The beginning of the carbon to burn is called the breaking
through of the flame, and when it is all burned, the drop of
the flame. The latter point is sharp and marked so that an
inexperienced eye can soon catch the point where the flame
drops. While a slight increase in phosphorus is shown, it
amounts only to the gain from concentration, i. e., the actual
weight of phosphorus is the same in the blown metal as in
the pig iron, whereas the weight of the latter is consider-
ably less (about 8 per cent.) than that of the pig iron; this
applies to sulphur also and there is usually a gain of both
phosphorus and sulphur corresponding to the loss in blow-
ing. This loss will depend mainly on the percentages of
carbon, silicon, and manganese in the iron ; the loss is not
only the actual amounts of these, but iron is always oxidized \
an increase of silicon calls for an increased amount of iron
in the slag, as the silicon in forming the double silicate of
iron and manganese takes up more iron, unless an unusual
amount of manganese is present, as ferrous oxide FeO and
manganous oxide MnO can replace each other to a large
extent. The combined percentages of these two oxides in
the slag amount to 30 or 35 per cent, in most cases, together
with from 60 to 65 per cent, of SiO^.

16. Tenii>erature In tlie Converter. — Silicon is the
great heat producer in the acid Bessemer process. The oxi-
dation of carbon and manganese produce considerable heat —
large quantities, in fact — but not enough for the reaction,
as is clearly shown by the fact that a decided decrease in the
percentage of silicon causes the metal to work cold. For-
merly 2 and even 3 per cent, of silicon was considered

18 MANUFACTURE OF STEEL § 34

necessary to furnish the requisite heat, but this amount has
been reduced so that the average metal going into the con-
verter to be blown contains .9 to 1 per cent. This decrease
has been due mainly to discontinuing the use of scrap and
to more rapid work throughout the process. It leads to a
great economy, as the loss is decreased not only by the les-
sened silicon, but by more than an equal amount of iron
taken up by the slag. In the case of cold heats, side blowing
is resorted to — the vessel is turned down to or approaching
a horizontal position, until some of the tuyeres are exposed
above the surface of the bath, and as the air is blown over
its surface iron is oxidized, its burning producing heat. It
is an expensive way to get temperature, but occasionally the
only way to get out, as heats are sometimes unavoidably too
low in silicon or blow cold from other causes, such as low
temperature of the metal from mixer or cupola, cold con-
verter, etc.

In using the higher silicon metal of former practice, the
vessels were always scrapped; i. e., steel scrap from the
rolling mill was thrown in during the progress of the blow.
The amount was determined by the blower, who signaled a
workman whose duty it was to throw in the weight required
to cool the metal sufficiently. It acted simply by absorbing
heat in melting, also by diluting the heat-forming elements
in the bath. The present practice is to turn steam in
with the blast. This method lessens the labor for handling
scrap; and further, in the development of the open-hearth
process it is more economical to use the scrap there. A
lower silicon mixture can also be run more safely.

17. Recarbonlzlng. — Recarbonizing is done in the
ladle; for low-carbon steels by the addition of heated ferro-
manganese, and for high-carbon and manganese steels by
using melted spiegeleisen or pig iron ; or, in the latter case,
the recarbonizer is frequently poured into the vessel. The
amounts necessary to furnish given percentages in the steel
are given, together with the loss, etc., under the heading
** Recarbonization.*'

§ 34 MANUFACTURE OF vSTEEL 19

18. Steel Jja<lle and Crane. — The ladle is of the ordi-
nary shape of riveted plate steel and is poured from the
bottom, as in the open-hearth. They are not usually
bricked up, but are lined up with 3 or 4 inches of ball
stuff consisting of ganister and clay. Patching is done
with the same, or, more commonly, with loam. Electric
traveling cranes have been installed for handling the iron to
the vessels, and the steel for casting, but are not considered
.so well adapted as the older swinging hydraulic cranes.
Generally, one crane pours the pig iron into the converter
and another handles the steel ladle and pours the steel.

19. Casting:, Etc. — This operation is common to both
the Bessemer and open-hearth processes, and is accomplished
in about the same manner. The older practice, and one
that was universal until within the past few years, was to
have a circular or a semicircular pit, with the steel crane in
the center and the molds placed around its circumference,
so that the crane could reach any part. The molds and
heats were made to correspond, so that a heat would give
an even number of ingots and avoid butts, which are either
inconvenient to handle or must be remelted as scrap. Just
as in the open-hearth process, or more correctly it was first
done in the Bessemer process, practically all plants cast the
ingots in molds carried on cars, two or three to each car.
This method avoids a pit, always a dirty part of the plant;
but economy is the controlling motive in such matters.
The molds are pushed to the stripper, usually in a separate
building, and removed from the molds, either the ingots
being left standing on the cars and the molds removed and
placed on other cars, or 'the mold and the ingot ar6 both
removed and the ingot pushed out on cars to be taken to the
heating furnaces — the soaking pit ox pit furnace,

20. General Arrangrement of Plant. — Fig. 6 shows a
common general arrangement of a Bessemer plant. Details
of arrangement vary greatly with the judgment of the
engineer designing the plant or limitations imposed by sur-
roundings. In the plan shown, the four cupolas ^ are placed

MANUFACTURE OF STEEL

§34

at one side, and at right angles to them the two convert-
ers b. Opposite the latter is the pulpit e^ from which the
blower directs the blowing operations and controls the

Pig. 6

cranes d for the iron and steel. Behind the cupolas are
shown two lifts, or elevators /, for raising the stock to the
cupola charging platform, and two molds r on a car, to be
served by the steel crane d^ are sh6wn.

THE BASIC BESSEMER PROCESS

21. Introductory. — The basic Bessemer process bears
the same relation to the acid Bessemer process as the basic
open-hearth process does to the acid open-hearth. Conver-
sion is accomplished in the same way as in the acid Bessemer
process — by blowing air through the molten iron — with the

§ 34 MANUFACTURE OF STEEL 21

essential difference that purification is effected by introdu*
cing a lime charge; the basic slag resulting requires the use
of a basic-lined vessel. It renders available for steel making
irons entirely too high in phosphorus for the acid Bessemer
process and also too high for economical use in the ordinary
basic open-hearth process.

The presence of phosphorus in the pig iron in the early
work of Bessemer seemed likely to render the pneumatic
process a failure, and Bessemer gave considerable time to the
removal of it. But on finding iron within the allowable limits
of phosphorus required to produce steel, and that much suit-
able pig iron was available, he abandoned his experiments.
As the process extended, low-phosphorus irons became
relatively scarce and dear in England and on the Continent,
so that many of the leading German and English metal-
lurgists gave their efforts to dephosphorizing, but without
success. While several accomplished it experimentally, no
practical results were reached until Sidney Gilchrist Thomas,
a young English metallurgist, achieved success in 1877-1878.
Associated with him was his cousin, Percy C. Gilchrist,
a steel- works chemist; and it was for a time known entirely
as the Thomas-Gilchrist process^ but later and at present as
the basic process. It is an accepted fact that in 1872
George J. Snelus, one of the leading English steel metal-
lurgists, discovered the means of dephosphorization by using
lime in a converter. He did not, however, carry his experi-
ments to final success, and the work of Thomas was carried
on independent of this, so that he is entitled to full credit
for originality. The principle of the basic process was first
applied to the Bessemer and afterwards to the open-hearth
process, the latter being now much the more important of
the two basic processes. Germany has made the greatest
development of the basic Bessemer process, mainly owing
to available pig irons better adapted to it than to the basic
open-hearth process. But two plants have been started in
America, and neither is now in operation.

Thomas worked at intervals for 7 years on the problem of
dephosphorization. He collected all the analytical and

22 MANUFACTURE OF STEEL § 34

technical data on the subject, and soon came to the conclusion
that in order to eliminate phosphorus a strong base should
be added with the charge, so as to retain the phosphorus
when oxidized and carry it off in the slag ; also, that this
condition demands either a basic-lined apparatus or one
not attacked by the basic slag formed. After experiment-
ing jfirst with crucibles and later with small converters lined
with every possible basic refractory material and using a
great number of alkaline and alkaline-earth salts, Thomas
finally settled on dolomite for the lining and lime for the
basic flux. To successfully make this lining required a long
trial with various admixtures and methods of treating the
dolomite. After thorough calcination of the dolomite at a
very high temperature, it is mixed with tar, molded into
bricks, and these again burned at a heat that will sinter
them.

23. Pig: Iron Used. — The essentials in the pig iron are
a low-silicon and high-phosphorus content. It was at first
thought that moderate percentages of the latter (under
1 per cent.) could be used to advantage, but later practice
demonstrated the necessity for 2 or 3 per cent, of phos-
phorus for the best results, as the oxidation of this element
furnishes the bulk of the heat, instead of the silicon, as in
the acid Bessemer process. Silicon could be almost, if not
entirely, dispensed with, but it is impossible to make pig
iron otherwise suitable (low enough in sulphur) without con-
siderable silicon. It should be below .5 per cent, and should
in no case exceed 1 per cent., the latter being too high for
an average mixture. The chief reason why low silicon is
imperative is on account of the lime used and the basic slag
required, so that the smallest amount of silicon possible
must be in the charge if a sufficiently basic slag is to be
produced without an excessive use of lime. Manganese
ranges in the practice of different works from .75 to 3 per
cent. ; from 1 to 2 per cent, may be taken as the usual limits.
The higher manganese is required to furnish some of the
heat required at the beginning of the blow — the low silicon

§34

MANUFACTURE OF STEEL

not giving enough heat at this stage, the manganese, being
oxidized immediately after the silicon, supplies the defi-
ciency. A further advantage is the desulphurizing ten-
dency of manganese, as basic Bessemer pig is apt to be high
in sulphur, owing to the low silicon required. Sulphur is
removed to a slightly greater extent than in the basic open-
hearth process, and may therefore be somewhat higher in
the pig metal to produce the same sulphur content in the
steel. It should not exceed .05 per cent, to make very low
sulphur steel, nor . 1 per cent, in any case. Carbon is some-
what lower than in ordinary pig iron, usually 3 to 3.5 per
cent. Owing to the low silicon, high manganese, and phos-
phorus (all of which promote this tendency), the carbon is
mostly combined, giving the pig a white or silvery-gray
fracture. The pig iron is either melted in cupolas or taken
directly from the blast furnaces, the same as in the acid
Bessemer process. Table V gives the usual limits of analysis
of pig irons.

TABIiE V

ANALYSES OF BASIC BESSEMER PIG HIONS

Works

Middlesbrough, England .

Kladno, Austria

Witkowitz, Austria

Horde, Germany

Creusot, France

Pottstown, Pennsylvania

Silicon.
Per Cent.

I.O to 1.3

1.2 to 1.3

.4 to .8

.2 to 1.2

1-3
below .5

Manganese
Per Cent.

.6 to 1.0
.3 to .5

1.0 to 1.4
.5 to 3.0

1.5 to 2.0
.8

Phosphorus.
Per Cent.

1.5 to 2. 75

1.5
.9 to 3.40

1.2 to 2.60

2.5 to 3.00
2.5 to 3.00

Sulphur.
Per Cent.

.050 to .12

.105
.080 to .13
.050 to . 10

.200
.020 to .05

23, Basic Converter — liining:. Bottom, Tuyeres. —

The converter is constructed, the same as the acid vessel, of
heavy plate steel mounted on trunnions so as to be rotated.
Owing to the large amount of slag and the lime charge, it
is from 50 to 60 per cent, larger than the acid converter for
the same iron charge. The usual capacity is from 6 to

24 MANUFACTURE OF STEEL § 34

15 tons of metal. As the converter requires much more
repairs than in the acid process, to run as continuously,
three vessels are generally installed so that two may be
available for use while the third is being relined. Another
method, originally proposed by Holley, is to have the entire
vessel removable, so that it may be taken away either by an
overhead crane or on a car and a freshly lined vessel, relined
and dried in a separate shop, substituted.

The lining is built up of basic bricks, made of lime,
dolomite, or magnesia, using a mortar of the same, mixed
with tar. More often, the lining is rammed in of the same
basic materials mixed with a little clay and about 10 per
cent, of anhydrous tar to give plasticity and act as a binder
for the dead-burned material. The usual thickness of the
lining is from 12 to 24 inches at the bottom and from 8 to

16 inches at the nose. Constant repairs are required
between heats, using the lining material for this purpose.

A converter averages about 100 (75 to 125) heats on a
lining, but occasionally gives out on the first heat or two.
The slag destroys the lining, especially at the nose, by build-
ing up on it when pouring at the end of the blow, and on
removing this, part of the lining is apt to come with it, or
the slag, by not being sufficiently basic, rapidly attacks the
lining.

The greatest difficulty in the early history of the basic
Bessemer process was experienced in making the bottom.
It is rammed up, similar to an acid bottom, of the same
material as the lining — either the basic ball stuff with tar,
or the basic brick and this together. It is rammed in layers
until a thickness of 20 to 26 inches is obtained.

The tuyeres are the ordinary clay ones used in the
acid process. Originally it was considered essential to use
tuyeres of basic material, as it was held that clay would flux
with the basic bottom ; but the clay tuyeres have proved to
be better and are much more economical. They are distrib-
uted over the bottom and the material is rammed around
them, as in the acid-process furnace. Instead of using clay
tuyeres, the bottom is sometimes rammed up around iron

§ 34 MANUFACTURE OF STEEL 26

pins about i inch in diameter, and on withdrawing them, the
holes left serve for the tuyeres. The number and size corre-
spond to those of the acid Bessemer process furnace. When
made up, the bottoms are dried and thoroughly burned before
being placed on the converter.

24. Blowing:. — The first part of the operation, or the
foreblow^ corresponds to the acid Bessemer process, when the
silicon, manganese, and carbon. are removed. The phos-
phorus is removed at a distinct stage and later termed the
afterblow. The blast is furnished from blowing engines,
but a higher pressure (from 25 to 35 pounds per square
inch) is required than in the acid process for the same
size heat, depending on the size of heats, the shape of the
vessel, etc. The burned lime is first introduced on the
bottom, generally previously heated, or coke or coal charged
with it and the blast slightly turned on to burn the latter
and heat the lime charge. The latter varies from 10 to
18 per cent, of the weight of metal and depends on the
amount of silicon and phosphorus in the metal, as well as on
the purity of the lime. The foreblow lasts from 10 to
12 minutes, and the afterblow about 5, or the entire time of
blowing averages from 15 to 18 minutes, although occasional
blows last much longer, owing to variations in the charge,
conditions of tuyeres, or other causes. If the metal is too
hot, scrap is added during the blow, the same as in the acid
process, to reduce the temperature. The metal should be
as hot as possible during the first part of the blow, as this
prevents slopping of the charge from the converter, but the
temperature should be reduced before pouring the steel so
as to give a sufficiently viscid slag to avoid rephosphoriza-
tion in the ladle. Too fluid or too hot a slag will allow
reactions to start in the ladle and some of the phosphorus to
be reduced from the slag and returned to the metal.

The loss in blowing depends mainly on the character of the
metal, and averages about 14 per cent, of the pig iron
charged, but may vary from 11 to 19 per cent. In addition
to the oxidation of the metalloids, iron is oxidized to form

26 MANUFACTURE OF STEEL § 34

the slag, the same as in the basic open-hearth process, the
amount depending on the percentage of silicon and manga-
nese in the pig — higher silicon requires more iron Fe^ as well
as calcium oxide, and higher manganese, less iron. In
general, the slags correspond to basic open-hearth slags,
but are higher in phosphorus, owing to the initial charge
being so much higher in phosphorus.

26. In blowing, the conditions are judged from the
character of the flame, as in the acid Bessemer process, for
the foreblow, and regulated accordingly, by varying the blast
or by addition of scrap. The afterblow, during which the
phosphorus is oxidized, is determined entirely by the volume
of air blown through the metal, and no attention is paid to
the flame or other indications, for turning down the vessel.
When the change comes, i. e., when the carbon flame drops
(carbon being practically all oxidized), the revolutions of
the blowing engines are shown by revolution counters
placed in the pulpit, showing the blower the volume of air
delivered. This is determined by experiment for different
percentages of phosphorus, but is approximately one-half
the length of the foreblow. In starting with a new mixture,
when the blower judges the phosphorus to be removed, the
vessel is turned down and a sample taken out with a test
spoon or ladle and poured into a small mold. This is
rapidly hammered out under a steam hammer, cooled
quickly, and broken, a record is made of the number of
blows required to break it, together with the character of
the fracture indicating the degree of dephosphorization :
if brittle and weak, the metal is cold short (high in phos-
phorus) ; a crystalline fracture light in color and having a
general appearance soon recognized, but not easily described,
^ows to the experienced eye whether phosphorus is low
enough, with almost the certainty of an analysis. This test
having been made, the blow is continued or the steel poured
as the result may indicate. After a number of blows the
preliminary test is discontinued and the volume of air alone
relied on for the dephosphorization, the analysis of previous

§ 34 MANUFACTURE OF STEEL 27

heats, reported before the succeeding heat is ** turned down,"
being closely followed at the same time to check any varia-
tions in the charge or blowing. As phosphorus is the main
heat producer, its percentage is regulated as the charge
blows hot or cold from a variation of the other constituents
— ^low initial temperature of the metal as it is poured into
the converter, the rapidity of working, the condition of the
bottom and the tuyeres, or other cause.

26. Oxidation of the !EleinentB. — As previously indi-
cated, the foreblow corresponds closely to the acid blow, the
elements being oxidized in the same order. Silicon is the
most readily oxidized and the first to burn, forming SiO^ —
which combines with iron and manganese to form ferrous
and manganous silicates. The other elements cannot be
oxidized to any great extent so long as any appreciable
amount of silicon exists in the bath, as the oxides of iron,
manganese, carbon, and phosphorus are all reduced by
silicon. In the acid Bessemer process the only exception to
the preceding (phosphorus not being affected) is that in the
case of a very hot working charge. Carbon may be oxi-
dized before all the silicon is, the affinity of the two for
oxygen being reversed at very high temperatures. This,
however, will rarely, if ever, occur in ordinary acid prac-
tice, and only with very high initial silicon ; in basic practice
it is much less likely, if not impossible, to occur, owing to
the basic conditions existing.

Manganese is the next most easily oxidized element, and
begins to burn before all the silicon is gone, but there must
be silica present to form manganous silicate or the silicon
will reduce any oxides of manganese as formed. Manganese
is not oxidized as rapidly as in the acid Bessemer process,,
as the lime present, by keeping the slag basic, makes less
demand for manganese. With low or moderate percentages
of manganese (say under 1 per cent.), the most of it will be
burned by the time the full carbon flame starts; with 2 or
3 per cent, of manganese, more will be burned with the car-
bon; in any case, the last tenth or two-tenths per cent.

28 MANUFACTURE OF STEEL § 34

remains nearly until the drop of the flame. Carbon is oxi-
dized to carbon monoxide CO^ and carbon dioxide CO^^ after
the silicon and part of the manganese are gone. Immedi-
ately after removal of the silicon, the carbon is oxidized
mostly to carbon monoxide ; and as the carbon in the charge
decreases, the percentage of carbon monoxide in the esca-
ping converter gases decreases ; or the less carbon that there
is in the bath, the more carbon dioxide is formed. Although
the combustion of the carbon produces a large amount of
heat, the bath gains a much smaller percentage of the heat
thus produced than from the combustion of the elements
yielding solid products, as the silicon, manganese, iron,
and phosphorus. This is owing to the carbon combustion
products escaping as gases {CO and CO^) and carrying out
the greater part of the heat produced by its combustion;
this is especially true of the carbon burned to carbon mon-
oxide, the bath retaining but little of this heat. This applies
equally to the acid or basic Bessemer processes.

Phosphorus is oxidized to phosphorus pentoxide P^O^ and
combined as calcium phosphate Ca^{PO^)^j or ferrous phos-
phate FeJ^PO^^^ but almost wholly as the former, though
some authorities believe the latter is present to a consider-
able extent. The phosphorus cannot be oxidized until the
other elements are completely removed It exists in the
iron as phosphide and is converted (during the afterblow)
to phosphate; the latter would be reduced by silicide of iron or
the carbides of iron or manganese to phosphide and returned
to the metal if silicon and carbon were present. The fact
that it cannot be oxidized while the other elements remain,
although a basic slag exists, also indicates that it forms
calcium phosphate entirely. Sulphur is removed to a con-
siderable extent in regular basic Bessemer practice, and if
the initial charge contains high manganese or if manganese
is added at the close of the afterblow, much more is remo'ved.
It is further effected by an overblow, i. e., continuing the
blow after the phosphorus is oxidized and at the expense of
oxidizing iron, the slag, rich in ferrous oxide (and, better,
manganous oxide), acting on the sulphur. In the acid

§34

MANUFACTURE OF STEEL

I .. *:

Iphur
rCen

1 OS

»o

\n Tf

»o

> O

3 a>

•

* •

•

■

C/3 CU

3 *J

phor
Cen

► o

!>•

1 NO

»o

1 M

O 4)

1 C4

P ^

. 4J

c c

•S<5

»o

•

1 Tf

> •

•

cd u

to «

o s,

■

(0 -«->
4) G
G 4>

) 00

•

• •

•

ilicon
rCen

> r*

»o

ITi

io O

•

1 O

1 •

•

■

•

^ £

— t

•
•

: B

a.
o

Xfl

M-4

' s

0 ««

c
o
y

• •-4

i §

t-l

c«

v»-

•
•

•

•
•
•

•

•
•
•

m
•
•

•
•
•

•»^

► 0)

1 CA
1 lO

CA
00

«s

c«

•<t

•

1 s

u O

• ^i^

• »"

*t-^

• ^m

• V4

• 9^

• ^4

i E

4->

r> 00

MANUFACTURE OF STEEL

§34

process the overblow is the blow after the removal of the
carbon; in either case an excessive oxidation of iron occurs;
it may be done in addition to the above, either to give some
heat or remove the last traces of carbon.

Table VI, which is taken from Wedding's ** Basic Bessemer
Process," shows the successive removal of the elements in
blowing.

27. Action of the Basic Fluxes. — The functions of the
basic lining — dolomite or lime — and the lime charge have
already been given, the latter to effect the dephosphorization
and the former that the necessary basic slag may be carried
so as not to destroy the vessel's lining. The lime, of course,
neutralizes the silica from the oxidation of the silicon in the
pig. The percentage of silicon mainly determines the amount
of lime to be used, but the phosphorus also must be taken
care of by the lime charge ; in practice, a considerable excess

TABI.E VII

Number

•

. c

CaO.
Per Cent.

•

MnO.
Per Cent.

FeO.
Per Cent.

CaS.
Per Cent.

12.07

11.74

55.94

5.37

2.48

6.08

1. 91

2. 88

12.77

16.92

47.87

6.75

4.80

5.94

2.87

.09

7.35

16.79

50.66

7.13

4.71

7.85

3.98

1.06

7.20

19.20

49.00

3.75

4.26

9.00

4.83

.92

of lime is used so as to be on the safe side, as a deficiency
will cause injury to or destruction of the lining as well as a
failure to effect dephosphorization. It is held that the cal-
cium phosphate can form only when the slag is saturated
with bases to a lower silicate and an excess of lime is present.
While phosphate of iron may be formed at first, it is gradu-
ally changed to calcium phosphate before the afterblow is
completed — the iron being displaced by the stronger base.
While there may be some question as to just how the differ-
ent compounds exist in the slag, it is reasonably safe to

§34 MANUFACTURE OF STEEL 31

assume that iron and manganese combine as ferrous and
manganous silicates, together with a part of the lime, to
form an extremely basic silicate; another portion of the
lime combines with the phosphorus; and the balance
remains free to give additional basicity to the slag; a
small amount forms calcium sulphide with part of the
sulphur present.

Table VII, which is taken from Wedding's ** Basic Besse-
mer Process," gives analyses of typical basic Bessemer
slags.

28. Owing to the large percentage of phosphoric acid in
the slag, it becomes a valuable by-product for fertilizing
purposes; it is largely used for this purpose in Europe,
where large quantities of it are produced. Many methods
have been proposed for preparing the slag, but it is now
ground in a ball mill to extreme fineness and applied in this
form. As the phosphate is insoluble, only a part of it is
available as plant food, and for this reason it is less valu-
able than soluble phosphates. In cases where the pig iron
is not high enough in phosphorus, the slags are returned to
the blast furnace and smelted over. The high percentage
of lime and magnesia, usually from 55 to 60 per cent., make
it of value as a flux, besides recovering the iron and manga-
nese contained. If limestone is expensive and the pig iron
is not made too high in phosphorus by using the slag in the
blast furnace, this is frequently the most economical means
for its disposal. This is governed by local conditions as to
cost of flux and its value as a fertilizer.

29. Tropenas Process. — This process is adapted to
making steel castings and is carried out in a special Besse-
mer vessel in which the blast of air is blown on top of
the metal instead of through it. From very early in its
history, top-blown or side-blown converters have been used
at different times in the development of the Bessemer proc-
ess. The ^Robert or Walrand converter is practically the
same as, and an earlier one than, the Tropenas. The latter

82 MANUFACTURE OF STEEL § 34

is shown in Fig. 7 (a) in vertical section, while Fig. 7 {3)
is a horizontal section through the lower wind box e. It
has an upper and lower wind box e and / on one side of
the vessel, from each of which horizontal tuyeres extend
through the side of the ves-
sel. They are of large
diameter, from H to
2 inches, and so placed that
the ends are always above
the bath. The upper row
of tuyeres d are placed
from 4 to 7 inches above the
lower c and are not used
until the metal is desilicon-
ized by the lower row and
the carbon flame starts,
when air is admitted to the
upper wind box to burn the
carbon monoxide formed
from the oxidation of the
carbon. The purpose of
this is to utilize the heat
of the carbon monoxide,
which is largely added to
the bath by radiation. The
blowing is stopped, as in
the ordinary converter,
when the carbon flame
disappears ; the recarbon-
izatiOn is made in the con-
verter and the metal f>oured
into a ladle for casting. It
is claimed that a much
hotter metal is obtained than by the usual practice, that
more delicate and intricate shapes can be cast from it,
and that the quality of the metal is improved. The ves-
sels are small — of 1 or 2 tons capacity. The advantages
are the cheaper installation than an open-hearth, as the

§ 34 MANUFACTURE OF STEEL 33

converter may be placed in any foundry with a cupola
to melt the pig iron. The pig metal used is from 2.5 to
3 per cent, silicon, .04 to .1 per cent, phosphorus, .03 to
.06 per cent, sulphur, .5 to 1.25 per cent, manganese, and
3.5 to 4.25 per cent, carbon; it is the high silicon that
gives the excessive temperature which permits pouring of
difficult castings. Owing to the high silicon, the loss in
blowing is excessive, reaching from 10 to 12.5 per cent.
The metal is recarbonized to the same composition as ordi-
nary castings. A low-blast pressure (3 or 4 pounds per
square inch) is used. For making light castings the process
is in use in many places in Europe and in ten or twelve
works in America with some twenty converters.

RECARBONIZATION

30. General Remarks. — The term recarbonization^ or
recarburizatiofiy is used to cover more than the mere adding
of carbon to the metal ; all the ordinary additions, as of man-
ganese and silicon in different alloys, are included. Perhaps
a more exact term than recarbonizer would be additions^ but
the former has the sanction of usage. Its purpose is to give
the metal the required properties as to strength and quality,
and also that it may be handled in the subsequent opera-
tions of casting, rolling, forging, etc. The metal in the
furnace or Bessemer converter after blowing in the condi-
tion of almost pure iron (generally containing but a few
hundredths of a per cent, of impurities), is worthless for
practical purposes and requires the various additions,
depending on the use to which the steel is to be put. The
recarbonizer is not only for the purpose of leaving certain
amounts of carbon, manganese, or silicon in the metal, but
also for removing objectionable compounds from it. These
are gases and oxides of iron or other elements. Hydrogen,
nitrogen, and carbon monoxide are commonly present in
steel, rendering it wild, i. e., causing violent ebullition of the
metal in the furnace, ladle, or molds. No theory as to
their introduction or elimination is fully accepted, their

34 MANUFACTURE OF STEEL § 34

effect being neutralized or their removal accomplished by
the additions. Oxygen, either as the free gas or in the form
of oxides — metallic or gaseous — is the chief cause of wild
steel, and the recarbonizer, acting as a deoxidizer, reduces
these. Silicon and aluminum act in the same way (but their
presence is not usually desired in the steel except in steel
castings or in special steels), and are mostly removed by
the action of the oxides in the bath.

31. The two general methods of recarbonization are, in
the furnace or converter and in the ladle. In the Bessemer,
process, it is nearly always done in the ladle; the recar-
bonizer is thrown in as the metal is poured, so as to insure
a better mixture. In the case of high-carbon or manganese
steel, or where molten pig iron is used to furnish the carbon,
it is frequently poured into the converter and a low blast
turned on long enough (a few seconds) to thoroughly mix it
with the metal, but not enough to cause any loss by the
blast.

In the open-hearth process, a part or all of the recar-
bonizer is added in the furnace ; if in part, the balance is
added in the ladle. In the case of large heats, the addition
is not heated whether thrown into the furnace or ladle, but
with small heats it sometimes is. In Bessemer practice, it
is either heated to redness for soft steel, or melted for rail
or other steel high in carbon and manganese, where more is
required. In general, when the recarbonizer is added to a
large bath it is not necessary to preheat it ; whereas with a
relatively small bath, as in the Bessemer or small open-hearth
furnaces, it must be heated or melted, according to the
amount to be used, so as not to appreciably lower the tem-
perature of the metal.

32. Recarbonizers. — The usual recarbonizers are :

(1) Ferromanganese, an alloy of manganese, iron, and carbon ;

(2) spiegeleiseUy the same as the preceding, except the
amount of manganese is much less; (3) ferrosilicon^ a very
high-silicon pig iron; (4) silicospicgil^ a very high-silicon

§34

MANUFACTURE OF STEEL

spiegeleisen, or a very high manganese ferrosilicon ; (5) car-
bidi of silicon (carborundum), an alloy of carbon and silicon
produced in the electric furnace.

The first four are made in a regular iron blast furnace
from properly selected ores and with special manipulation
of the furnace and special burden. The fuel consumption is
excessively high — usually two to three times that required
for pig iron — and the output greatly reduced, which, with
the higher cost of the ores, explains their high price — from
three to five times the cost of pig iron.

Table VIII gives analyses showing the usual range of com-
position of recarbonizing alloys.

TABIiE Tin

Ferromanganese. .

Spiegeleisen

Ferrosilicon

Silicospiegel

Silicon carbide

Carbon.
Per Cent.

Man-
fifanese.

Per Cent.

Silicon.
Per Cent.

Phos-
phorus.

Per Cent.

Sulphur.
Per Cent.

6.5107.0

78 to 82

.3 to 2

.20 to .3

.01

4.6105.0

12 to 20

.3 to 3

.15 to .3

.01

I.5t0 2.5

I to 4

lo.o to 20

.10 to .3

.01

2.0 to 4.0

15 to 20

10.0 to 15

.15 to .3

.01

62.0

trace

350

none

Iron.
Per Cent.

10.0 to 15
72.0 to 80
75.0 to 85
65.0 to 75

1-5

Alloys above 20 per cent, of manganese are usually classed
as ferromanganese ; below this, spiegeleisen. The standard
ferromanganese is 80 per cent, manganese, and little else
is used until we come to spiegeleisen, 20 per cent. In ferro-
manganese, between 20 and 80 per cent, of manganese is
sometimes met with, but seldom at the present time. Owing
to the conditions of manufacture, sulphur is never beyond a
few thousandths of a per cent. Phosphorus is governed by
the ores, but should not exceed .3 per cent., and this gives
from .002 to .003 per cent, of phosphorus in the steel.

Carbide of silicon has only been used a few years, and
finds considerable application as a source of carbon and
silicon, especially in the manufacture of steel castings. It
is always added in the ladle, and produces a violent reaction

36 MANUFACTURE OP STEEL § 34

if any considerable quantity is used. Its advantag:es are
that a smaller amount is required for the same increase in
silicon or carbon, and that the temperature of the steel is
somewhat increased instead of lowered, as in the case of the
metallic alloys. This increase is due to the heat developed
by the combustion of silicon and carbon, but mainly to the
fact that the decomposition of the compound releases a large
amount of heat. Owing to the very high temperature at
which the carbide is formed (6,500** F.), a large amount of
energy is stored up as latent heat.

33. Recarbonlzation In the Furnace. — This method
has certain advantages and drawbacks, as compared with
making the addition in the ladle. With the heat ready to
tap, the recarbonizer is thrown into the furnace and is
allowed a few minutes to melt and mix with the bath. It is
claimed for this practice that the manganese is more thor-'
oughly mixed, the bath is more thoroughly deoxidized, the
temperature of the metal is not lowered, and the metal is
quieter and less likely to irregularities in casting. Against
this claim there is said to be a greater loss of manganese,
silicon, etc., than when the recarbonizer is added to the
ladle. The loss will depend on conditions of the bath and
time in the furnace before tapping. A bath containing a
large amount of oxygen will oxidize more manganese or
silicon than one nearly free from solid or gaseous oxides. A
higher temperature, with other conditions constant, will
also cause a greater loss. In making high-manganese steel,
the actual weight lost is more, but the percentage of loss
remains fairly constant, but increases somewhat for differ-
ent amounts in the finished steel, i. e., when other condi-
tions affecting the loss of manganese remain the same. In
the acid open-hearth practice, the manganese is generally
about all burned out of the bath, while in basic practice it
will seldom be reduced below .1 per cent, and may be as high
as .25 per cent., depending on the amount in the melting
stock — mainly the pig iron. The recarbonizing loss in the
acid is somewhat greater than in basic work, but for both it

§ 34 MANUFACTURE OF STEEL 37

may be taken at 30 or 40 per cent, of the manganese added
No hard-and-fast rule can be given for th;e amount of ferro-
manganese to be added for a given percentage in the steel,
as the conditions stated above affect the loss, and the melter
is guided by the conditions of the bath, the time in furnace,
and the results from previous heats. The following examples
from actual practice show the amount used :

1. A 32,000-pound charge in an acid-process furnace was
taken, and from .36 to .4 per cent, of manganese was wanted
in the steel. To effect this, 250 pounds of 80-per-cent.
ferromanganese (which equals 200 pounds of metallic man-
ganese) was added to the charge in the furnace. Assuming
a loss of 40 per cent, of manganese, we have .60 X 200, or
120 pounds of metallic manganese to be absorbed by the
bath; 120 -^ 32,000 = .00375, or .375 per cent, of manga-
nese in steel by calculation, allowing a loss of 40 per cent. ;
the analyses showed from .36 to .4 per cent, of manganese
on a large number of heats, showing that the allowance of
40 per cent, loss was correct in this case.

2. With the same charge, from .28 to .3 per cent, of
manganese was wanted. 175 pounds of 80-per-cent. ferro-
manganese, equal to 140 pounds of metallic manganese, was
used. Taking the loss at 40 per cent., leaves 60 per cent.,
or 140 X. 60 = 84 pounds of manganese to be absorbed.
84 -T- 32,000 = .0026, or .26 per cent, of manganese (calcu-
lated percentage); the steel analyzed from .28 to .3 per
cent, manganese. Taking .29 per cent, as an average, then
32,000 pounds X .0029 = 93 pounds of manganese in the
steel. 140 pounds (the total manganese added) less 93 pounds
(the amount actually in the steel) leaves 47 pounds, or 34 per
cent, of manganese lost.

As stated, the loss of manganese in basic practice is less
than in acid ; the following heats illustrate this :

3. The heat is 75,000 pounds; .38 to .42 per cent, of
manganese is wanted in the steel ; 400 pounds of 80-per-cent.
ferromanganese, equivalent to 320 pounds of metallic man-
ganese, was added. In this case allowance must be made for

MANUFACTURE OF STEEL

§34

.1 per cent, of residual mangainese ih the .bath, so that there
is required from the ferromanganese 75,000 pounds X .003,
or 225 pounds of manganese. 320 pounds of manganese
added less 225 pounds absorbed leaves 95 pounds, or 30 per
cent, of manganese lost.

4. This heat also is 75,000 pounds, but from .58 to .62 per
cent, of manganese is wanted. 650 pounds of 80-per-cent.
ferromanganese, equivalent to 520 pounds of metallic man-
ganese, was added. In this case the bath contained .15 per
cent, of residual manganese; .60 per cent, less .15 per
cent, equals .45 per cent, of manganese required from the
ferromanganese, or 338 pounds. 520 pounds added less
338 pounds in the steel, leaves 182 pounds, or 35 per cent, of
manganese lost.

Table IX shows the manganese additions and losses for
the four heats just given.

TABIiE IX

Process Used

Weight of Charge.
Pounds

Pounds of
Manganese Added

Equivalent in
Percentage of Heat

Manganese in
the steel

Manganese Lost

•

c
U

Per Cent.

From
Ferroman-
ganese

Per Cent.

Heat

Per Cent.

of Amount

Added

I
2

3
4

acid
acid
basic
basic

32,000
32,000
75,000
75,000

140
200
320
520

.44
.63

■43
.69

.29

.38
.40
.60

.29
.38
.30

.45

.16
.25
.13
.24

34
40

Numbers 3 and 4 had, respectively, .1 and. 15 per cent, of
manganese left in the bath ; figures are given to the second
place, or the nearest whole number. The larger the amount
of manganese added, the greater is the loss when other con-
ditions remain uniform. This is shown in the table, as well
as the greater loss for acid heats. It would be useless to
attempt a definite statement on these points or as to the
amount of manganese to be used for a given percentage in

§34

MANUFACTURE OF STEEL

the steel. To go further, it may be stated that the ordinary
loss of manganese when added in the furnace is from 30 to
40 per cent., which may be increased or decreased by varia-
tions in the practice, melting conditions, etc.

34. Recarbonization In the liadle. — In the Bessemer
process the recarbonization is done in the ladle entirely for
soft steel and almost entirely for high-carbon steel. In
the acid or the basic open-hearth practice, many steel
makers prefer to make all the addition in the furnace, while
many others add a part of the ferromanganese in the fur-
nace and a part in the ladle, generally about half in each ;
the latter, while not universal, is the more general practice;
a few add the entire amount in the ladle.

35. Ix)S6 of Mangranese. — With manganese, the only
Advantage of recarbonizing in the ladle is the economy, as
the loss is less and may be taken at from 15 to 30 per cent.,
or from 10 to 15 per cent, less than in the furnace. The fer-
romanganese is not exposed to the action of the flame and
much less to the slag, and the action of the metal must be
less vigorous in the ladle than in the furnace; all of which

TABIiE X

I
2

3
4

•

•o

tM "O

Si 90

o ■<

O TJ

« a

•73 S
C 9i

s£

0 e«

PU bo

acid

6o,ooo

260

acid

6o,ooo

325

basic

8o,ooo

304

basic

8o,ooo

375

Q M

0 O

e« bo

> 5

O* 4>

.43
.54

.38

.47

Manganese in
the Steel

1.1

.35

.42

.36

.48

4) S g «
U O g «

fe ^ ^^ SI

.35

.42

.28

.36

Manganese Lost

OU 0

.08

.12

.10
.II

« o »
o

19
26

26
24

go to explain the smaller loss. For medium- and large-size
open-hearth heats the recarbonizer is not usually heated,
but thrown into the ladle, so as to mix with the stream of
metal. Occasionally it will be heated to redness, or always

40 MANUFACTURE OF STEEL § 34

so when the amount is excessive in the case of high-manga-
nese or silicon steel. Table X shows a record of heats recar-
bonized in the ladle.

Heats 3 and 4 retained, respectively, .08 and .12 per cent,
of manganese in the bath.

36. lioss of Silicon. — In steel castings, or other steel
requiring an addition of silicon, it may be added in the fur-
nace or ladle in the form of any of the silicon alloys men-
tioned under ** Recarbonizers." In the Bessemer process it is
usually added in the ladle, but may be thrown into the con-
verter with a gentle blast on for a few seconds, as it is oxi-
dized very quickly. This is sometimes done with cold heats
or casting heats wanted excessively hot. The loss of silicon
in the vessel or furnace depends mainly on the time in the
bath, and is subject to wider variations than is manganese, as
silicon is more readily attacked than manganese. It may be
taken approximately at 50 per cent, of the amount added,
and working under uniform conditions, is readily controlled,
but variations in the practice of different plants give differ-
ent results. In the basic process the loss is higher than in
the acid, owing to the slag having a greater affinity for 5/(?,.
If it is added in the ladle, the loss is much less (from 25 to
40 per cent.). In using carbide of silicon, the loss is from 50
to 60 per cent, of the silicon.

31. Control of Carbon. — So far no account has been
taken of carbon. A certain amount is added with the ferro-
manganese, the use of an 80-per-cent. ferromanganese rais-
ing the manganese about 12 to 15 times as much as the
carbon ; while spiegeleisen will give 3 or 4 times, owing to
the lower ratio of manganese to carbon. While 80-per-cent.
ferromanganese may be used in recarbonizing all grades of
steel, its distinctive use is for soft steel, where the desired
manganese content can be given without raising the carbon
appreciably. To get the carbon wanted, when above very
soft steel, molten pig metal was formerly poured into the con-
verter, and in the open-hearth the heat was caught coming

§ 34 MANUFACTURE OP STEEL 41

down, 1. e., at about the desired carbon, or pig iron added
to the bath. In Bessemer practice, this method is largely
followed yet for rails and other high -carbon steel. The
steel is blown down soft, i. e., practically all the carbon
burned out (down to .05 or .08 per cent, of carbon), and the
necessary amount of melted spiegeleisen and pig iron poured
in and the converter turned over to mix the addition, or the
latter may be poured into the ladle at the same time as the
blown metal and the mixing accomplished there.

38. An alternate method is the adding of solid carbon in
the form of crushed coke or anthracite. This is the Darby
method of recarbonizing, but the name of the inventor
is seldom mentioned in connection with it, although it
was only developed in 1888. The coke or coal is weighed
into ordinary paper sacks of a weight that each sackful will
give .01 or .02 percent, of carbon to the steel; these are
then thrown into the ladle as the steel is poured. About
one-half the carbon is absorbed by the steel, this depending
somewhat on the temperature — a very hot heat taking up
more than a cold one. In the open-hearth the practice
varies between (1) tapping when the carbon has been boiled
down a few points (hundredths of a per cent.) below that
wanted in the finished steel, and (2) boiling the heat down to
about .1 per cent, of carbon and recarbonizing back in the
ladle with coke or anthracite. In making high-carbon steel
in the open-hearth, say from .5 to 1 per cent, of carbon,
the second method is not practicable for all the carbon, and
the heat is always caught coming down and only a part or
none added, as required. Usually, it is not attempted to
make steel over .40 per cent, carbon by adding all the carbon
in the ladle when the heat has been blown (in the Bessemer)
or boiled down (in the open-hearth) soft, i. e., to about .1 per
cent, carbon, which means recarbonizing about .3 per cent.
The heat may be tapped, of course, between that wanted
in the steel and .1 per cent, carbon, and whatever coke or
anthracite is needed is added in the ladle. Table XI shows
the additions in the ladle for carbon, loss, etc.

MANUFACTURE OF STEEL

§34

TABIiB XI

No.

Test

I
2

5
6

Process Used

open-hearth
open-hearth
open-hearth
open-hearth
Bessemer
Bessemer

Weight of
Charge.

Pounds

6o,ooo
6o,ooo
iio.ooo
110,000
22,000
22,000

Carbon

Required

Bath.

in Steel.

Per Cent.

.10

.20

.15

.24

.10

.18

.75

.90

.08

.35

.08

.22

Coke or

Anthracite

Added.

Pounds

100

150

325

140

Carbon
Absorbed.

Per Cent.

56
47
52
52
46

The amount added will be governed somewhat by the per-
centage of manganese in the steel and whether furnished by
ferromanganese or spiegeleisen, the latter adding more car-
bon for the same amount of manganese than ferromanga-
nese. In the above table, .02 per cent, carbon is allowed
for the ferromanganese used. The coal or coke contained
85 per cent, of carbon.

TABIiE XII

Weight of

Wanted

in Steel

Amount Added to Heat

Heat.

Carbon.

Manganese.

Spiegeleisen.

Pig Iron.

Per Cent.

Pounds

10,000

•30

.65

400

150

16,000

•45

.70

650

900

22,000

.40

•75

1,000

800

22,000

•50

.90

1,200

1,150

32,000

•45

.80

1,500

32,000

•50

I 00

•

1,800

1,400

39, Table XII shows the amount of liquid pig iron or
spiegeleisen used to recarbonize — generally used in Besse-
mer plants making rails or other high-carbon steel regularly.

§ 34 MANUFACTURE OF STEEL 43

In this case spiegeleisen is preferred as a source of manga-
nese, owing to the higher carbon per unit of manganese.
About 90 per cent, of the carbon is absorbed, as it is already
in solution and only has to mix with the larger body of blown
metal and no chemical action or absorption has to take
place ; the loss of manganese is less than if added cold or
only heated to redness. The carbon is taken at about
.08 per cent, in blown metal; spiegeleisen, 5 per cent,
carbon, 20 per cent, manganese; pig iron, 3.5 per cent,
carbon.

It should be remembered that the weights of the recar-
bonizing additions are subject to change, as results ob-
tained are higher or lower than wanted, and working con-
ditions— temperature, pig iron, blowing, and manner in
which additions are made — affect loss of carbon and
manganese,

40. Use of Aluminum. — Metallic aluminum is very
generally used for quieting basic steel and, to some extent,
acid steel also. It acts as a deoxidizer and belongs under
recarbonizers in the general use of the term. 100 parts of
oxygen combines with 87.5 parts of silicon; 100 parts of oxy-
gen combines with 112 parts of aluminum; 100 parts of
oxygen combines with 344 parts of manganese. While a
given amount of silicon will combine with more oxygen than
the same amount of aluminum, the latter has a much greater
affinity for oxygen under the conditions and is therefore the
more powerful deoxidizer; but it is the least apt to remain
in the steel if oxides or free oxygen are present. It is always
added in the ladle or molds, from 2 to 5 ounces per ton being
used. If much above this is added, it causes too rapid solidi-
fication and defects from piping and cracking. In addition
to removing gases and making the steel quiet, it has the
property of rapidly permeating the entire mass of the steel,
which causes other elements to alloy more uniformly, pre-
venting or lessening segregation ; it gives sounder ingot tops,
thus lessening the loss as scrap; and it also slightly increases
the strength of the steel.

44 MANUFACTURE OP STEEL § 34

THE CRUCIBLE PROCESS

41. General Remarks. — The crucible process is the
oldest and simplest of the three principal ones, both in appa-
ratus employed and in manipulation. It consists essentially
in melting the stock in a crucible set in a bed of coke or
anthracite on the bottom of a vertical or shaft furnace. It
may be broadly defined as melting an iron either high in
carbon, requiring no recarbonizer, or melting one low in
carbon, demanding recarbonization. A number of melting
holes are constructed together and connected by flues to a
stack, thus forming the furnace. At present, it may be said
that practically all crucible-steel melting furnaces are of
the Siemens regenerative gas type.

45J. Crucible Furnace. — The furnace contains from
two to twenty holes, taking four or six crucibles each. Each
hole has its own gas and air regenerators, so that it is prac-
tically a separate furnace, but all the holes of a furnace have
a common stack and main flues. Sometimes separate valves
for controlling the gas and air to each set of checkers are
provided, but more commonly the one set, c for gas and d
for air, as shown in Fig. 8, are put in for the entire furnace.
The dampers / are placed in the air and gas flues to the
stack. Fig. 8 {a) shows a cross-section of the furnace and
pair of regenerators, a' for air and g* for gas on each side
of a melting hole; / and/' are gas and air ports, respect-
ively; / are flues under checkers leading to the stack.
Fig. 8 (b) is a longitudinal section on the line A By showing
four melting holes, the gas and air valves a and g^ respect-
ively, and stack s. Each hole o has two or three (the latter
number in the figure) movable arched coverings, or bungs j\
of firebrick held by clamps, which are lifted by hooks sus-
pended on a trolley for charging and drawing the crucibles.
Six or eight inches of coke dust is placed on the bottom of
each melting hole, in the center of which is a hole A, so that
if a crucible breaks, the steel runs into the vault v run-
ning the length of the furnace ; this is cleaned out at the

. §34 MANUFACTURE OF STEEL 45

end of each week. The melting holes are separated by
cross walls k.

43. Cmcibles. — These are of clay and graphite. Clay
crucibles are quite commonly used in England and in Europe

generally. In America, graphite crucibles are exclusively
used. They cost more than clay, but last longer and are
stronger, thus allowing larger ones to be used. The clay
crucible is held to be tougher at a steel-melting heat,

46 MANUFACTURE OP STEEL § 34

but IS very weak when cold, the walls not standing the
sudden contraction as well as the graphite, and if used

over, must be returned to the furnace

-. — ^ ^ as soon as the charge is poured. They

' ■ ■ are more apt to break in the fur-

I 1^ ^^ ^1 nace also, causing a greater loss of

steel from this source. Crucibles ordi-
^ ( ^ narily hold from 80 to 125 pounds;

the walls are from 1 to 1^ inches thick.
The dimensions of a 100-pound cru-
cible are shown in Fig. 9. The life
of a graphite crucible is from 3 to

Fig. 9 ^ r-

8 heats usually, and frequently only
one. This depends on a number of circumstances: the
quality of the crucible, depending on the materials and
manufacture; the kind of steel melted; they having a longer
life with high-carbon steel; whether plunged into a very
hot furnace or brought up more gradually; and the care
and skill of the puller out in drawing.

44, Materials of Whlcli Crucibles Are Ma<le. — Cru-
cibles are made from a mixture of about 50 per cent, graphite,
35 to 40 per cent, clay, and the balance sand. This varies
with the practice of the manufacturer, the quality of the
materials, and somewhat with the results desired. Graphite
is the well-known mineral quite widely distributed. It is a
form of carbon. The best is the Ceylon graphite, but
much native graphite is used. It is found in many parts
of the United States, especially in Wisconsin and New York.
The Ceylon product is the most valuable, not only owing to
higher purity, but the laminated, or elastic, fibrous struc-
ture serves to bind the matrix of clay more firmly than the
amorphous graphite, which is held to give much inferior
results. It should be ground rather fine, as if left too coarse
the crucible may become porous; if too fine, the walls are too
dense and it does not expand or contract so quickly when
exposed to sudden heating or cooling, and cracking results;
heat is conducted more slowly also.

§34

MANUFACTURE OF STEEL

Table XIII gives the analyses of several samples of graph-
ite. The more impure graphite is concentrated by dressing,
consisting in air floating or treating by wet methods. The
impurities accompanying it are generally iron pyrites,
gneiss, or limestone.

TABIiB XIII

Source

Carbon.
Per Cent.

Ceylon 99. 68

Canada 97-63

German (raw) 53- 80

German (dressed) 89. 20

Ash.
Per Cent.

.21
1.78

Volatile
Matter.

Per Cent.

.11
•59

Up to the present time most of the clay u.sed in crucible
making has been imported from Europe, although some
New Jersey, but more particularly Missouri and Colorado,
clays have had a limited use. There appears no good
reason from composition and properties why many of our
native clays should not be used. As in many other indus-
trial enterprises, the manufacturer is influenced by preju-
dice and tradition. Kaolin is used to give the proper
fusibility to the mixture. Good crucible clay must be strong
and plastic as well as refractory. It is a silicate of aluminum
with small percentages of other bases and a large amount
of combined water, which gives the plasticity. The objec-
tionable constituents are oxide of iron, alkalies, and alkaline
earths, as they all reduce the refractory qualities.

Table XIV gives analyses of standard clays and kaolins
used in crucible making.

Of these samples, number 1 was the famous Crown brand
from Kluengenberg, Germany; 2 and 3, the Rhenish clay,
Germany; 4, the Meisner clay, Germany; 5, kaolin from
Staten Island; and 6, kaolin from Brandy wine, Pennsylvania.

MANUFACTURE OF STEEL

§34

The sand used is the ordinary fire, or silica, sand having
from 95 to 99 per cent, of silica with small amounts of
alumina, alkaline earths, or combined water. Oxide of iron
and alkalies are the most detrimental constituents, as they
lower the fusing point, if present beyond a small amount.

TABiiB xrr

Number of
Sample

•

CaO.
Per Cent.

MgO.
Per Cent.

0 ^^.

if ^ S

c gfou

^ 0

Moisture at
Per Cent.

Alkalies.
Per Cent.

59.20

25.40

I. 71

.52

.42

8.34

4.14

46.99

30.04

2.14

•59

.55

11.69

4.18

3.00

.81

45.53

36.15

.25

.50

10.48

5.82

1.75

54.51

31-42

.68

.04

.43

12.37

.55

85.24

11.20

.72

.27

•45

1.26

65.80

22.09

1.58

.27

.32

7.09

.19

2.35

45. Manufacture of Crucibles. — This includes the
four processes of mixing, molding, drying, and burning. The
ingredients are mixed by paste mixers and clay- working
machinery to a thoroughly homogeneous mass, water being
added to temper it properly ; the batch when ready for mold-
ing contains about 22 per cent, of water.

Molding was formerly a hand operation, but at the present
time is mostly done by various machines, jigs, presses, etc.
The shapes, being quite simple, are readily formed by molds
and shapers with the machines.

The drying of the green crucibles must be done very care-
fully, as the shrinkage is so great it may so distort the
crucible as to render it useless or crack it. The average
shrinkage is about 5 per cent, from the water used in mixing,
but mostly from the combined water of the clay.

The burning requires the same care as the drying, and
may be considered the final stage of the latter. It is done
in some of the types of pottery kilns, the fuel being wood,
gas, or coal ; if the last, the sulphur must not be excessive, or

§ 34 MANUFACTURE OF STEEL 49

the crucibles may be injuriously affected by its absorption.
They are usually in the kiln 5 or 6 days — being fired at a
gradually increasing heat for about 3, and the kiln or oven
allowed to cool slowly for 2 or 3 days. The temperature
reaches about 750° or 800° C. (say 1,400° or 1,500° F.), and is
only required to take the entire shrinkage out of the clay.
Coming from the kiln, the crucibles have a color ranging
from a gray to a dark drab, depending partly on the tem-
perature, but much more on the character of the flame main-
tained— if oxidizing to any extent the graphite will be
burned from the surface, giving the light color of the clay
body ; if a reducing flame was kept during the firing, the
crucibles will be darker colored, depending on the amount
of graphite oxidized. Kilns are constructed to admit as
little air as possible in excess of that required to effect com-
bustion of the fuel, so as to reduce the oxidizing effect on
the crucibles.

46. Crucible Chargre. — The materials for making cru-
cible steel are chiefly puddled iron and wrought iron and
steel scrap, together with the necessary amount of carbon,
usually charcoal ; manganese, as ferromanganese or oxide of
manganese ; or other additions. Blister steel, made by the
cementation process of soaking iron bars with carbon in a
converting furnace at a red heat, was originally used, and is
even yet to a small extent in A^^crica, and quite largely in
the original home of crucible-steel making, Sheffield, Eng-
land. It is also held that the very highest grade of crucible
steel can only be produced from blister steel made from
the purest Swedish irons, even though other iron or soft
steel may be produced of the same composition. It is
impossible to give any satisfactory reason for this, and it
has been attributed to prejudice and usage handed down
through many years. As those best competent to judge,
and to whose interest it would be to use other stock, insist
that the higher priced Swedish irons give better tool steel,
the fact can only be accepted, with the statement that our
methods of examination are not perfect enough — whether

50 MANUFACTURE OF STEEL § 34

chemical, physical, or microscopical — to show us the distinc-
tions or combinations that give this superiority. Compara-
tively little of the latter stock is used in this country in
crucible melting, but the fact is of sufficient importance to
be brought out prominently. The materials are usually very
low in sulphur and phosphorus, as none is removed ; it is
an acid process, although basic crucibles have been used
together with a basic slag to effect purification, but this has
scarcely been more than an experiment, and has no promise
of commercial value or technical importance.

The crucible is carefully filled with the stock while cold
and then inserted into the melting hole. The practice in
England is to first place the crucible in the furnace, and
when it has been heated somewhat, to introduce the charge
by means of a sheet-iron funnel. As clay crucibles are gen-
erally used there, this allows a preliminary test before
charging and defective ones may be thrown out.

Packing the cold crucible outside the furnace allows the
stock to be more carefully placed, the larger pieces and the
charcoal for carbonizing and any oxide of manganese or
ferromanganese used on the bottom ; the smaller and closer
fitting pieces are packed on top and likely serve to keep any
gases from penetrating into the metal; also oxygen from
the charcoal, lessening the loss of the latter. The crucible
is then set in the melting hole by means of tongs. In the
regenerative furnace they are set directly on the coke-
breeze covering the bottom, or in a shaft furnace they are
partially embedded in the glowing anthracite or coke.

47. Melting. — This is generally divided into the sub-
divisions of melting and killings or dead-melting. With the
crucible in the melting hole, a cover is put on it to keep
out the gases. The temperature of the furnace is gradually
brought up, if a gas furnace, by adjusting the gas and air
supply, and draft, if necessary, to give the proper melting
conditions. In the case of a coke hole, the solid fuel is piled
up around the crucible to its top; if coke, it must be replen-
ished two or three times during melting; anthracite, owing

§ 3-t MANUFACTURE OF STEEL 51

to its compact structure, does not have to be renewed for
one melting. When the melter judges the charge about
melted, the covers are removed and the contents of the
crucibles examined to see their condition. The trained eye
of the melter at once recognizes the condition of the steel,
whether completely melted or if the temperature is too high
or too low, and adjusts the furnace conditions accordingly.
Sometimes the eye alone is depended on for temperature, or
a light iron rod is introduced and stirred around in the
metal, as in the open-hearth proce.ss. If the metal is very
hot, little or no steel adheres to the rod or it may be melted
off sharply at the end; if cold, the metal is sluggish and
pasty, building up on the rod and adhering to it when with-
drawn.

48. Killing:, or Dead-Melting:. — This is simply holding
the steel at a melting temperature until a change occurs that
gives sound ingots or castings. The change is doubtless the
simple one of the gases being boiled out of solution in the
metal. This action is probably assisted by the absorption
of silicon reduced by carbon from the SiO^ of the crucible
walls. The effect of killing is also held to be that the silicon
absorbed increases the power of the metal to hold gas in
solution, enabling it to retain while solidifying any gas in
the molten steel. This last explanation, while given by high
authority, cannot be held to be proved or better grounded
than the simpler one of boiling out any gas in solution. The
latter is commonly accepted by practical steel metallurgists.

The melting time is usually from 2^ to 3 hours. This
depends on a number of conditions, but principally
(1) whether hard or soft steel is being made — soft (low-
carbon) steel may require f hour longer for melting than
hard (high-carbon) steel, as the wrought-iron or other very
low-carbon stock of the former melts at a much higher
temperature than high-carbon stock; (2) the presence of
manganese as oxide or in the metallic state shortens the
time; (3) the furnace and its manipulation; (4) to a less
extent than the preceding, the character of the stock aside

63 MANUFACTURE OP STEEL § 34

from its composition, size of pieces, packing, etc. ; the
crucible — thickness of walls, their composition, etc.

There is no absolute line between the melting proper and
the killing, as this is interpreted by the judgment of the
melter, and the two periods overlap to some extent. Kill-
ing usually takes from ^ to 1 hour — it may be longer or
shorter, depending on conditions. Other conditions being
the same, the hotter the furnace, the shorter is the time
required for the killing; the purer the steel, the longer is
the time required, doubtless owing to the higher tempera-
ture necessary to bring the desired condition, which may
be merely the question of an ebullition to get the gases out;
the lower the charge is in phosphorus, sulphur, silicon,
manganese, or carbon, the more heat is required to give the
same degree of boil. The entire time in the furnace from
charging to drawing is generally from 2^^ to 3^ hours, so
that three charges are usually melted each 12-hour shift, or
turn, some little time being required between drawing and
a subsequent charging for teeming and some fixing of the
coke bottom in most cases.

49. Teeming, or Pouring:. — This operation is accom-
plished by lifting the crucibles out of the melting holes by
suitable tongs, picking them up with another pair, and pour-
ing into the molds for ingots or castings. It is done almost
universally by manual labor and is some of the hardest and
hottest work of steel manufacture, as the ** pullerout " must
straddle the melting hole while withdrawing the crucible.
Cranes with special tongs have been used to some extent
for charging and drawing; their use is not yet common,
however, but will undoubtedly become so. The molds are
of a size to hold the contents of one or several crucibles, in
the case of larger ingots, or a number of crucibles poured
into one casting. Crucible-steel ingots of 90 tons have been
made at the Krupp Works, Germany, requiring some
2,000 hundred-pound crucibles. In such a case, the most
careful selection of the stock is essential to insure uniformity
of the ingot ; and perfect organization and discipline of the

§ 34 MANUFACTURE OF STEEL 53

large number of men, so as to have the teeming done with
sufficient promptness. Such ingots are made only there,
and are used for armor plate. Many others of large size are
made for high-grade forgings, such as engine shafts, pro-
peller shafts, and other marine forgings, also guns and
gun forgings. In America, such materials requiring large
masses of steel are always made of open-hearth steel. In
the ordinary crucible shop, making tool steel mainly, the
ingots are about 3 or 4^ inches square and the weight of
one or more crucibles full. The molds are split lengthwise
and held together by rings keyed on. Before teeming and
while separated they are smoked by burning rosin, coal tar,
or a smoky gas flame ; this acts as a mold wash and gives a
better surfaced ingot.

60, The loss in melting is very low — the least of any
steel process — usually being from 1 to 3 per cent, of the
weight of metal charged. The cost of melting is the highest
of any process, approximately from 15 to 16 per ton, or
from three to five times the labor cost in the Bessemer or
open-hearth process. The fuel consumption is high com-
pared with the latter, about 1 pound of coal as producer gas
per pound of steel, or about 15,000 cubic feet of natural gas
per ton of steel melted; approximately, three times the
amount required for open-hearth melting. The above fac-
tors, together with the limited output and the higher priced
melting stock that must be used, explains the comparatively
limited field of crucible steel, which is restricted to purposes
where the first cost of the steel can be ignored — mainly tools,
fine springs, saws, files, fine machinery parts, etc.

51. Superiority of Crucible Steel. — While no fully
satisfactory reason has been given for the superiority of
crucible over other grades of steel of like composition, the
causes generally given are: (1) The purer stock melted;
(2) as the crucible is covered during the melting, the gases
from the fire have very little chance to be absorbed by the
metal.

54 MANUFACTURE OF STEEL § 34

It seems safe to say that to the conditions of melting are
principally due the finer quality of crucible steel. In regard
to the purer stock, there can be no direct comparison with
Bessemer or open-hearth steel, as it is impossible to make
either from the usual crucible stock without the use of other
materials. But in the crucible process the melting recep-
tacle is closed and all gases are largely kept from the steel,
whereas in the Bessemer process the air is blown through
the molten metal, exposing it to the oxygen and nitrogen of
the blast, the solid and gaseous products of combustion,
some of which are undoubtedly absorbed, affecting the prop-
erties of the steel. In the open-hearth process much the
same conditions may be found, except that the gas for oxi-
dation plays over the surface of the bath.

52. The widest ranges of composition are possible, and
obtained regularly by varying the mixture. Carbon may be
from .1 to 2.25 per cent., but as practically all crucible steel is
used for tools or purposes requiring similar grades, we may
restrict the carbon between .4 and 1.5 per cent, as covering
the bulk of the product. Manganese varies between .1 and
.75 per cent., but most grades are below .6 per cent; sili-
con, between a few hundredths of a per cent, and .2 per cent.,
although considerable is made above this, not including
silicon steel. Sulphur and phosphorus are each kept below
.02 per cent, as a rule, but for less exacting purposes this is
frequently exceeded, but seldom above .05 percent, of either
element is allowed. In the highest grades, where the purest
Swedish melting stock is used, sulphur and phosphorus may
not exceed .01 per cent. The effects of impurities are
decidedly more marked in high-carbon steel than in low; the
metal seems to be more sensitive, and the same amount of
sulphur, phosphorus, or silicon influences the properties
more.

53. Crucible steel is divided into different grades, accord-
ing to temper or carbon content, one temper generally mean-
ing .1 per cent, carbon. In determining the grades of the
steel, the ingots are broken, or topped, and graded by the

§34

MANUFACTURE OF STEEL

fracture. Sometimes color carbon tests are made, but most
crucible shops use the fracture for grading purposes, and an
experienced eye seldom misses, the carbon more than .05 per
cent. While no sharp subdivisions exist as to the uses to
which different grades of crucible steel are put, the follow-
ing shows them in a general way:

Steel of from .5 to .75 per cent, carbon is used for batter-
ing tools, hot work, dull-edge cutting tools, etc.

That from .75 to 1 per cent, carbon is used for dies, axes,
knives, drills, and similar purposes.

That from 1 to 1.5 per cent, carbon is used for razors,
lathe tools, gravers' tools, little drills, etc.

The best all-around tool steel is between .9 and 1.1 per
cent, carbon, and is capable of being adapted to a wider
range of uses than any other grade. Between .9 and 1 per
cent, carbon iron is saturated with carbon, giving the best
results in tools and highest strength.

Table XV shows the analyses of various crucible steels
and purpose used for.

TABIiE XV

Use

Sledges, battering

tools, etc

Hot work shear

knives, etc

Drills, reamers, dies,

etc

Chisels.knives, lathe

tools, etc

Razor steel

Dies, graving tools,

etc

Cutting tools, etc.

(self-hardening). .
Krupp armor

C
0U

.65

.85

1. 00
1.30

1.30

.94

.28

£ S

a ^

.21

.20

.18

.26
.22

.16

1.50
.32

c
o
o

.210

.180

.210

.200
.200

.140
.160

•055

I 2

.022

.020

.015

.010
.006

.014

.015
.016

Phosphorus, i
Per Cent.

Tungsten.
Per Cent.

Nickel.
Per Cent.

Chromium.
Per Cent.

.020

.015

.014

.010

.009

.012

.OT2

3.40

.015

3.('>o

1.75

56 MANUFACTURE OF STEEL § 34

While the above are analyses of samples for the uses indi-
cated, the composition of steel for the same purpose will
vary within cpnsiderable limits, depending on the practice
of the steel maker, but more especially on that of the user,
as to tempering and the exact use to which it is put, speed
of machine, if a cutting tool or machine part, and character
of work to be done by drills, tools, etc. Different manufac-
turers will produce the same quality of steel or give the
same properties, by varying the percentages of carbon, sili-
con, or manganese. In general, manganese, silicon, sulphur,
and phosphorus fluctuate but little, carbon being the vari-
able element that gives the desired temper. In Table XV
only low manganese is given among carbon steels. Occa-
sionally crucible steel is made with manganese from .3 to
1 per cent, (not considering alloy steels), but practically all
of it contains .2 to .3 per cent, of manganese. Sulphur and
phosphorus sometimes exceed tlie amount shown in the
commoner grades, and in extra-special grades both are reg-
ularly kept at from .005 to .008 per cent.

AliliOY STEEIiS

54. General Remarks, — By alloy steels are meant steels
that owe their special properties to the presence of other
elements than carbon. The carbon, however, generally
plays an important part in these special or alloy steels, while
in the ordinary or carbon tool steels their properties are due
almost wholly to the carbon present.

55. Tun^^eu Steel. — In some respects the most impor-
tant of these is self-hardening steel, sometimes known as
musket steel. It is a steel that hardens without quenching
in water or other liquid, when previously heated to the right
temperature, in this case to about a medium orange color.
It owes this property to tungsten and is sometimes called
tungsten or air -quenched steel. It is sometimes quenched in
a blast of air, to give greater hardness than if allowed to
cool in quiet air. It may be made much harder still by

§ 34 MANUFACTURE OP STEEL 57

quenching in oil or water, but the strains set up within
it are sufficient to overbalance its cohesive power and crack-
ing results, or it is so brittle as to crumble when used. It
is so hard when air quenched that it cannot be machined or
touched by the hardest carbon steel. By annealing it at
about a bright-orange heat for from 24 to 36 hours, and cool-
ing very slowly by covering it in the furnace with hot sand
or ashes it will be annealed so that it can be machined quite
readily. It becomes brittle at the full steel-working tem-
perature or below an orange color. It can be worked readily
between an orange and bright-orange heat. As its use is
restricted to cutting and machine tools, they are forged
as nearly to the desired shape and size as possible and are
then ground to the exact dimensions. It is not as strong
as good high-carbon steel ; while it can be made hard enough
by water quenching to cut chilled cast iron, the cutting tool
will not stand up to the work, the edge crumbling down.
Its chief advantage is that its temper is retained at relatively
high temperatures almost to a visible red, where a plain
carbon steel would have its temper lowered so as to be use-
less until retempered ; this enables it to do more work at
high speeds, allowing lathes, planers, boring mills, etc. to
be run much faster or heavier cuts to be taken, which means
great economy in the machine shop.

56. Tungsten itself is not believed to be directly the
hardener, but indirectly through its action on the carbon
and manganese. Steel has been made low in carbon and
manganese, with 3 per cent, of tungsten that would not
temper even when quenched in water; raising the carbon
but leaving the manganese low, it would harden like ordi-
nary carbon steel, but not in air; the addition of 2^ or 3 per
cent, of manganese gave the usual self -hardening steel,
showing that manganese and carbon are essential for the
tungsten to perform its part in air quenching, or that it acts
indirectly by its effect on carbon and manganese.

In regard to the hardness being retained at quite high
temperatures, the tungsten has been called **the mordant

58 MANUFACTURE OP STEEL § 34

that holds the carbon in solution*'; with plain carbon steel,
working at a high speed or severe duty, and the tool get-
ting hot, the carbon in effect comes out of solution and the
temper is lost; whereas, with the tungsten holding the
carbon in solution the temper is retained.

The percentage of tungsten may vary from . 1 to 10 per
cent., the latter being very unusual and difficult to obtain.
It is usually from 3 to 5 per cent. It may be introduced
in the crucible in the form of ferrotungsten (an iron-
tungsten-manganese alloy), or as the mineral wolframite,
tungstic oxide WO^ associated with more or less iron and
manganese. This is readily reduced by the carbon of the
charcoal or the crucible walls: WO^ -f- 4C = IF + \C0,
Both methods are used in practice. The carbon is gov-
erned by the temper desired; manganese is always over
1 per cent, and may exceed 3, usually 1.5 or 2.5 per cent. ;
silicon, sulphur, and phosphorus are the same as in car-
bon steel.

57. Mangfanese Steel. — Steel containing much above
1.25 per Cent, of manganese is nearly as brittle and unwork-
able as spiegeleisen until it reaches from 5.5 to 6.5 per cent,
of manganese, when it improves so that a tough product is
obtained — between 6.5 and 20 per cent, of manganese.
From 7 to 14 per cent, of manganese gives the best results.
This is perhaps the most unique alloy met with in steel,
possessing both hardness and toughness beyond that of any
other. It is so hard that no steel tool will touch it, yet so
tough that castings, forgings, etc. made of it may be bent
and hammered like the softest of mild steel. These two quali-
ties are directly antagonistic when either is present to an
extreme degree in all iron alloys known up to this time. The
maximum strength is obtained at about 13 or 14 per cent, of
manganese, and the composition of the greater part made is
from 12 to 14 per cent. The steel is necessarily high in
carbon from the fact that the ferromanganese used is high
in carbon. The carbon is about one-twelfth of the manga^
nese; the latter may vary, however, from 10 to 15 times the

§ 34 MANUFACTURE OF STEEL 59

percentage of carbon depending on the grade of ferroman-
ganese used.

58. The steel is water quenched to secure the extreme
hardness and toughness; it is sometimes so in its natural
condition, i. e., as cast, but quenching always improves it.
This is another peculiarity of the metal, as in all other steel
an increase in hardness means an increase in brittleness — a
decrease of ductility and elasticity. Its hardness is not so
much added to by quenching as its ductility. The steel is
practically non-magnetic under the strongest influences. Its
use is necessarily restricted to parts not requiring machin-
ing— castings and forgings, mainly, that do not require fin-
ishing to extremely exact sizes. It works readily at a red
heat. Its principal uses are for the jaws or working parts of
crushing and grinding machinery ; cheeks and plates of rock
crushers, edge mills, etc. ; car wheels, axles, and tires to a
limited extent ; in general, where strength with great hard-
ness or ductility, or both, are required. One of the more
recent uses is in the manufacture of safes and vaults, a pur-
pose to which it seems especially suited when the construc-
tion difficulties are overcome. It may be made in the
crucible, but as quite large masses are produced, the open-
hearth is the more suitable apparatus, and therefore is
always used. Owing to the large amount of manganese,
solid castings are readily produced ; the metal is extremely
fluid, allowing small and light castings to be made. Its
shrinkage is excessive, about f inch to the foot, thus adding
to casting difficulties. It was originated by R. A. Hadfield,
of Sheffield, England, and is commonly known as Hadfield's
manganese steel.

59. Nickel Steel, — This steel is used chiefly for armor
plate, but has a large use besides in forgings and castings.
It raises the strength about 50 per cent, over that of ordi-
nary steel of the same carbon content ; it also increases the
elasticity and ductility. The amount present is usually
from 3 to 4 per cent. It has been used to some extent for

60 MANUFACTURE OF STEEL § 34

car axles and boiler steel for very high pressures, but cannot
be said to be fully accepted for either purpose. Engine and
propeller shafts are largely made of it or other forgings or
castings requiring a particularly strong and ductile steel. It
is made almost entirely by the open-hearth process, but may
be made by either the Bessemer or crucible processes. The
nickel is added in the form of metallic nickel or ferronickel,
which is charged with the rest of the stock, practically no
loss occurring in melting. The steel works readily hot or
cold, forges easily, and machines harder than carbon steel.

60. Chrome Steel. — Chrome steel is used somewhat for
tools, but mainly for giving very hard surfaces and to resist
severe shocks. Its chief use is for armor plate and projec-
tiles; also very hard dies, mortars for crushing very hard
materials, etc. For tools, it is manufactured in the crucible ;
but for armor plate in the open-hearth furnace (except pos-
sibly Krupp uses the crucible). The chromium is added as
ferrochrome; if in the open-hearth, after the desired carbon
has been reached, as chromium oxidizes easily and the loss
is heavy. The amount present is usually from 1.5 to 2 per
cent.

•

61. Silicon Steel. — Silicon steel has been made con-
taining 1 or 2 percent, of silicon, but it has no extended use.
It gives a very hard steel, but workable hot with difficulty —
hot short. It is made by the crucible process.

62. Molybdenum Steel. — This steel is made to some
extent for special uses and possesses properties somewhat
similar to tungsten steel, but is tougher. It is used for
some high-grade saws and a very few other purposes. It is
alloyed in amounts ranging from a few tenths of a per cent,
up to about 3 per cent.

63. Other Alloy Steels. — Aluminum, copper, and tita-
nium steels have been made experimentally, but are of no
use as yet.

MANUFACTURE OF STEEL

(PART 8)

STEEL CASTINGS

!• General Bemarks. — The manufacture of steel cast-
ings is an important branch of the industry, both techni-
cally and commercially. Casting steels are produced by
exactly the same methods and apparatus as other grades
of steel, similar stock being used, in either the acid or basic
open-hearth, Bessemer, or crucible processes. As a matter
of fact, the bulk of steel castings are made by the open-
hearth process, although both the other processes contribute
some. With the advance in engineering construction of all
kinds, ordinary iron castings cannot be given the strength
required; consequently, steel is being used for purposes
where high duty is required. It is largely used for gear-
wheels of all kinds, engine frames and parts, locomotive
driving-wheel centers, in electric and ship construction to
some extent, where forgings were formerly used, rolling-
mill and other heavy machinery, and, in fact, steel castings
may be substituted for iron castings of any description
where strength is an important factor. The cost of the
steel castings is necessarily greater, or it would supplant
gray iron even further than it has. In addition to increased
strength, parts may be made much lighter than from iron
for the same strength, thus making the weight of a finished
machine or structure much less, which is in many cases a
consideration of scarcely less importance than strength.

§35

For notice of copyright, see page immediately following the title page.

2 MANUFACTURE OP STEEL § 35

The basic open-hearth steel is used to some and an
increasing extent, but the acid open-hearth steel is the most
used. This is due to earlier troubles with the former proc-
ess, and makers have not had the confidence to use it,
especially when the spoiling of the steel means the further
loss of the foundry labor of molding, etc. The regular Bes-
semer steel is not used in any shop on steel castings exclu-
sively, but some plants making ingots make occasional cash-
ing heats. Several modified Bessemer processes have been
used in regular casting practice, one, the Tropenas, which
has met with much success, has already been described.

!8. Solidity. — In making castings, it is essential that the
steel lies comparatively dead in the molds, with little action,
otherwise the product will be more or less honeycombed
with blowholes, caused by the escaping gases. To over-
come this, the special knowledge and art of the maker of
steel castings are necessary. It is accomplished by the use
of deoxidants, or deoxidizers," which remove the gases while
the steel is molten, or increase the power of the metal for
holding them in solution.

Solidity is further due to a riser or sink head made on top
of the runner, or gate, so that it is above the casting, and as
the latter cools and contracts metal flows in from the sink
head and fills up the shrinkage cavity. The weight of the
sink head depends on the size and character of the cast-
ing. It may amount to 50 per cent, of the weight of the
latter, but is usually from 15 to 30 per cent. As its func-
tion is to supply molten metal to the contracting casting, it
must be large enough to remain open until the casting sets,
and also have enough liquid steel to supply the demands due
to shrinkage of the latter. The deoxidizers, silicon, alumi-
num, and manganese, remove the gases or increase the
solvent power of the steel for them. While their action is
not absolutely understood, they produce solidity by either
or both these actions. All steel in the melted state has in
solution gases, and its power to hold them so is largely
dependent on the temperature. Killing in the crucible

§ 35 MANUFACTURE OF STEEL 3

removes them (possibly with the aid of silicon), and solid
ingots or castings are produced. This effect is reached in
open-hearth or Bessemer castings by the use of silicon or man-
ganese in the form of some of the recarbonizing alloys or addi-
tions given under ** Recarbonizers"; or by adding metallic
aluminum, all of which come under the general head of recar-
bonizers or deoxidizers. An excessive amount of these can-
not be used, or the metal will be made brittle from the over-
dose or possibly from retaining too much of the gases; yet it
will be perfectly solid and free from blowholes. The latter
may not lessen the strength and toughness of castings to
the extent their presence would indicate, but in parts to be
machined or to have finished surfaces, their presence is
entirely unallowable. Carbide of silicon is used in some
steel foundries as the sole source of silicon and part of the
carbon; in others, silicospiegel for silicon and manganese,
or ferrosilicon for the silicon, and spiegeleisen or ferro-
manganese for the manganese.

3. Composition of Casting: Steel. — The composition of
the steel depends, as in other grades, on the use to which
the same is to be put. For very soft castings, where great
toughness and ductility are required, but not high tensile
strength, the carbon may be as low as .12 per cent. ; where
stiffness and great strength are wanted and ductility is of
less importance, carbon may be as high as .8 per cent. For
ordinary purposes and covering castings for most uses, the
carbon is from .2 to .5 per cent. The amount of silicon will
vary with the carbon, as a rule, from .1 to .4 per cent. — the
low-carbon steel having the less, and the harder (high-
carbon) the more, silicon. The usual range is from .2 to
.3. per cent.

The amount of manganese present is usually .5 to .8 per
cent., but it may be outside these limits. Some castings
are made with from 1 to 1.25 per cent, of manganese, and
are air quenched to toughen them; i. e., heated to a cherry
red and allowed to cool in the air. The amount of phos-
phorus may reach the usual Bessemer steel limit of .1 per

MANUFACTURE OF STEEL

§35

cent., but the best castings should not exceed .04 per cent.,
which is readily attained in basic practice, but in the acid
requires the use of higher-priced stock. Phosphorus is held
to produce brittleness under shock, and is therefore espe-
cially objectionable in castings subject to sudden strain or
shock.

Sulphur is of less importance in castings than in most
other grades of steel, as its influence is felt mainly in work-
ing at a red heat, and does not greatly affect the cold steel
when present in moderate amounts. Its usual range is
from .025 to .05 per cent., and should not exceed the latter
very much. Aluminum is added frequently as a solidifier
(deoxidizer), equivalent to from .02 to .03 per cent, (from 4
to 10 ounces to the ton of steel), but this is mainly oxidized
to Al^O^ in the slag, and the small amount in the steel can-
not be accurately determined* As in all ordinary steels,
carbon is the principal strengthener, manganese, silicon,
and sometimes alummum give solidity and freedom from
blowholes.

Nickel-steel castings are made to a limited extent where
greater strength and toughness *is wanted than is given by
plain carbon castings. It is used for pinions on heavy
rolling mills or for parts subject to sudden and severe
shock. The nickel is usually from 3 to 4 per cent, in
such steel.

Table I gives the analyses of some steel castings.

TABIiE I

Kind of Casting

Machinery castings. . .
Machinery castings. . .

Rolls

Pinions

Carbon.
Per Cent.

Manga-
nese.

Per Cent.

Silicon.
Per Cent.

Sulphur.
Per Cent.

.18

.30

.28

.032

.24

.60

.30

.040

.48

.45

.31

.036

.75

.80

.28

.040

.26

.45

.27

.056

.44

.74

.33

.045

Phos-
phoms.

Per Cent.

■

.082

.045
.032

.050

.060

.092

§ 35 MANUFACTURE OF STEEL 6

BEFECTS IN" STEEIi

4. Segrregratlon. — Unfortunately for the metallurgist
and user, large masses of steel are never absolutely homo-
geneous, and frequently wide variations are shown between
different parts of the same ingot. This difference in the
composition, or the tendency of certain elements to separate
out, is known as segregation. Occasionally it is so serious
as to render a part of the ingot unfit for use, but generally,
when proper care has been exercised in making and handling
the steel, its effects are not dangerous. With other condi-
tions the same, the larger the ingot, the greater is the segre-
gation. With very heavy ingots for armor plate, forgings,
or where great homogeneity and reliability are required, a
portion of the top is cut off for scrap or to be used for
inferior purposes.

The causes of segregation are fairly well understood ; it is
due mainly to the lower melting points of the iron carbides,
phosphides, and sulphides. As the metal freezes, these,
by remaining fluid at lower temperatures, are squeezed out
and collect in the part of the ingot last to solidify, which is
usually the upper central part, approximately the upper
fourth or fifth of the ingot. It occurs without any regular-
ity and the laws governing it are not understood. In gen-
eral, the greater the percentage of metalloids, the greater is
the liability to segregation and the more serious it will be.
If the steel in the ingot could be instantly solidified, with-
out otherwise injuring its properties, segregation would be
avoided ; so that slow cooling favors the separation of the
impurities ; and as their specific gravity is less, they have a
tendency when once formed to rise through the body of
metal.

5. The term segregation should be confined to those irreg-
ularities occurring after pouring into ingots or castings, as
distinguished from irregularities in the furnace or ladle.
The latter may be due to careless melting, or addition of
recarbonizers in such a way as not to be uniformly distrib-
uted throughout the metal; while evils of this kind have

6 MANUFACTURE OF STEEL § 35

been charged to segregation, it is well established that a
thoroughly uniform metal is generally gotten in the ladle
and there is little excuse for variation there. The same
cannot be said of the steel after it has been poured, as the
conditions under which segregation takes place are only
partially under the control of the metallurgist. The condi-
tion favoring homogeneity is that the steel remains molten
the least possible time permissible. If made to solidify too
quickly, as bad or worse consequences follow — cracking, the
formation of excessive blowholes, and piping. Casting at
excessively high temperatures or in very large masses are the
principal causes of segregation, and keeping both within rea-
sonable limits is the chief remedy for it. Both act by keeping
the steel longer in the liquid state, allowing more favorable
opportunities for the compounds of lower melting points to
separate out, i. e., mainly the carbides, phosphides, and sul-
phides ; manganese and silicon segregating to a less extent.
There is no rule or law yet known that controls the order or
extent to which the different elements segregate; but in
most cases it occurs as follows : Carbon, phosphorus, sulphur,
silicon, and manganese — both as to liability of its taking
place and the extent of it. There are many exceptions, but
generally it takes place as above. When excessive segrega-
tion of one element is found, others are to be looked for with
it, but this does not always occur. The use of aluminum,
by lessening the time the steel remains fluid in the
molds and causing it to solidify more evenly, diminishes
the evil.

6. In Tables II, III, and IV, examples are given of
some extreme cases. It must not be assumed that all steel
segregates seriously because no. examples of uniformity are
given. While all masses of large size vary somewhat and
absolute homogeneity is never expected, yet for practical
purf)oses steel may be assumed as being uniform, the many
exceptions either proving the rule or are to be explained by
special circumstances in the manufacture, chiefly casting
temperature and mass.

X —

§35

MANUFACTURE OF STEEL

O
QD

« c
c «)

s •

r*»NO »nO »nO "(fco inr^tn^r^ooo ^O^C^

8 o

-oii

0 C

M CO r» r^ >0

« O O O 0

CO w « r^r^»nQmt<:

ooooooooo

tfi ^ tn O
xn <n f^

o o o

I 8

■§• u

o o

•^ <-i O CD

M l-l •-• O

cor».2>2w%M>p^C4C«i^7C«ci

ooooo^oooooooo

5 c

Oooco oooco moo r^o ^cntn»nOOGOco

c
o

■*■*

^ 6
•3 o

ii E

•a o

o.'S -
215

a 8.

a
o

MANUFACTURE OF STEEL

§35

Table II shows sections through the ingot at the points
indicated, the top being about one-fifth of the distance from
the upper end of the ingot, or usual point of the greatest
segregation. Drillings were taken from the center of each
section.

TABIiS HI

BXAMPIaES OF CARBON 8EGREOATIOK

Top of

' Ingot

Bottom of Ingot

Number of

Outside.

Center.

Outside.

Center.

Ingots

Per Cent

Per Cent.

Per Cent

Per Cent.

Carbon

•37

•55

•50

•51

.50

.60

•50

•55

.55

.60

•50

•55

•63.

•45

.60

•50

•56

.42

.62

.45

.62

.40

.60

•50

•52

•55

Average

•50

• .5«

.48

.56

7. The variation is not entirely from top to bottom, but
also from outside to center — a shell chilling next the iron
mold first and the interior of the ingot remaining fluid, the
carbides, phosphides, etc., owing to their lower melting
points, are pushed out of the solidifying mass and enmeshed
in the gradually freezing steel. Table III illustrates this
for carbon, and when other elements segregate, their varia-
tion corresponds to carbon as a rule, but is usually less in
amount. The eight ingots in this table were from the same
heat, the regular ladle analysis showing .56 per cent, of
carbon.

§35

MANUFACTURE OF STEEL

TABIiE IT

SEOKBGATION OF STEEL CASTINGS

Description

o o

Broken steel roll, center

Broken steel roll, outside

Broken steel roll, end opp. break. . .

Broken steel roll, center

Broken steel roll, outside

Broken steel roll, end

Steel pinions, center

Steel pinions, outside

1.25
.90

.44
.65
.44
.19

.17
.26

.048
.030

.035

.331
.070

.063

.060

.050

a O

CO £

.060

.048
.043
.165
.046
.036
.056
.050

• S °

-go

C3 C
O V

CO ©

.58

.15

.55

.13

.52

.14

.85

.33

.54

.27

.52

.24

.45

.27

.44

.27

The foregoing castings weighed from 4,000 to 6,000 pounds
and afforded opportunities for segregation similar to large
ingots. In steel castings of medium and small size, segre-
gation is practically absent, as the mass is liquid a much
shorter time. It is generally less in castings, as the metal
is partially killed with silicon or aluminum, so that the freez-
ing interval is less.

8. Bloiivholes. — Blowholes are small cavities, usually
spherical in shape, formed in the ingot as the steel solidifies,
and are caused by bubbles of gas unable to escape through
the frozen mass. They may be due to some Extent to air
drawn down mechanically by the stream of metal while
pouring, but are generally accepted as coming from gases
either formed or escaping from the solution as the metal
sets in the mold. The principal gases are nitrogen and
hydrogen, but carbon monoxide is considered by some
authorities as playing an important part. Blowholes in low-
carbon steel cannot be prevented and do not cause injury to
the steel, as the inner surfaces of the cavities cannot oxidize
and are readily welded together by subsequent rolling or
forging. The purer the steel, other things being the same,
the more blowholes will be formed; high carbon, silicon, or

10 MANUFACTURE OF STEEL § 35

manganese usually causes the steel to lie quiet and be free
from blowholes. Dead-melting decreases the number of
blowholes, crucible steel being almost free from them ; any
addition causing the steel to lie quiet (kill or deaden it) will
decrease them. Blowholes are not to be regarded as alto-
gether objectionable, but rather as a necessary condition,
especially in the soft and medium grades of steel, and their
removal or prevention may be harmful. If a steel ingot be
broken, there will be found a solid skin, usually from ^ to
1 inch thick around the outside, depending mainly on tem-
perature of casting; with excessively hot steel it will be
very thin, and thicker with steel at normal casting tempera-
ture. Next to this skin are the blowholes, or honeycomby
extending around the ingot; they may spread well into the
middle, depending on the kind of steel and temperature in
pouring. As stated above, their volume will be greater
with soft steel. If brought too near the surface by very hot
steel, the skin is so thin that in reheating and rolling this is
removed or rolled into the honeycomb or blowholes, expo-
sing these on the surfaces of plates, a serious defect in the
latter, known as pitting from the small holes, or pits.

9. Pipes. — Pipes are shrinkage cavities in the upper cen-
tral part of ingots, formed after the outside has solidified.
The exact relation between blowholes and pipes cannot be
explained, but in general steel that does not form blowholes,
pipes more or less, and vice versa. As examples of this,
crucible steel is free from blowholes, but pipes more or less
deeply ; high-carbon or silicon steel exhibits the same ten-
dency; also, conditions in the same steel that lessen the
tendency to form blowholes generally increase the liability
to pipe; e. g., the addition of silicon or aluminum for quiet-
ing steel lessens the former, but induces piping, and this
may be quite marked even in soft steel, if an excessive
amount of silicon or aluminum is added. As a rule, the
fewer and smaller the blowholes, the greater the piping.
Extremes of casting temperature — either too hot or too cold
— increase both blowholes and piping.

8 35 MANUFACTURE OF STEEL 11

10. Prevention of the Formation of Pipes and Bloiiv-
lioles. — The precautions mentioned above — regulation of
the temperature mainly and certain additions — are the only
ones observed to control or prevent these two conditions in
ordinary practice. Many means have been tried, but the
only one used to any extent is the Whitworth system of
liquid compression, in which the steel is cast in strongly
reinforced molds or cylinders and while still fluid subjected
to a pressure reaching 1,500 pounds per square inch from a
powerful hydraulic press. This prevents both blowholes and
piping by producing perfectly solid ingots, at the same time
giving more uniform composition. Its use is limited, and
mainly, if not wholly, restricted to very large ingots for heavy
forgings, such as marine shafts, large guns, etc.

EFFECTS OF THE USUAIi ELEMENTS PRESENT IN

STEEL

11. General Remarks. — Only those elements com-
monly found in ordinary commercial steels will be con-
sidered here, all reference to special or alloy steels being
omitted. The constituents affecting the properties, and
those usually present in ordinary carbon steel, are carbon,
manganese, sulphur, phosphorus, silicon, and oxide of iron;
copper and nickel, being present in considerable steel, will
be included. While each element has its own distinctive
effect, it is frequently difficult or impossible to determine
just what this is in given steels, as the effect will be so
modified by the amount of one or more of the others present
or the almost endless combinations of different percentages
of the elements. Conditions in the making and subsequent
treatment in rolling, hammering, cooling, etc. mask or
exaggerate the influence of given amounts. There are,
however, certain well-defined effects for the different ele-
ments, and these will be given as generally accepted by
metallurgists.

la MANUFACTURE OP STEEL § 35

12. Carbon. — This is by far the most important of the
elements in steel. It combines in all proportions up to
about 2 per cent., but seldom exceeds 1 per cent., except in
tool or special steels (ferromanganese may contain 7, or
pure iron combine with 4.5 per cent.). It is readily absorbed
at or above a red heat and the metal does not have to be
liquid; manganese increases the affinity of iron for carbon.
In common steel the carbon is present as combined carbon,
though small amounts of graphite may occasionally be pres-
ent. Carbon increases the strength and hardness, but
decreases the ductility. Strength is increased up to .9 or
1 per cent, carbon; above this it diminishes; the melting
point of steel is lowered by carbon ; the nearer we approach
pure iron, the higher is the melting point. An increase of
strength and a loss of ductility and elasticity go together
with carbon steel.

13. Maiifiranese. — Manganese increases the strength and
ductility of steel, but its chief function is the effect it has
on other elements, mainly oxygen or oxides and sulphur,
acting as an antidote for red shortness — brittleness at a red
heat. Manganese alloys are used to recarbonize and
remove oxygen from the bath, although some of the latter
always remains, and the residual manganese neutralizes its
effect and that of sulphur. Sulpl\ur and phosphorus tend
to produce coarse crystallization, and manganese seems to
prevent this, giving a fine-grained fracture. It increases
the rolling qualities or hot working of any kind, i. e., gives
/lot ductility; it also allows steel to be heated hotter without
injury. Steel with very low manganese will crack in rolling
or forging along the edges, whereas the same metal with
higher manganese will usually work satisfactorily. While
manganese is not a panacea for bad steel, nor will it cover
up the effects of improper working or too high impurities,
it is the most essential addition in correcting necessary
evils — e. g. , the presence of sulphur and oxygen or oxides.
In soft steel, manganese ranges from .3 to .6 per cent. ; in
hard and medium steels, rails, forgings, etc., from .4tol per

§ 35 MANUFACTURE OF STEEL 13

cent. While no definite rule exists as to sulphur and man-
ganese, approximately 8 to 10 times as much manganese as
sulphur is allowed.

14. Sulphur. — The effect of sulphur is felt when work-
ing at a red heat, for with it the metal cracks and tears and
welds much less readily. The remedy was given in discuss-
ing the effects of manganese on steel, or rather a corrective,
as this is notably a case where ** prevention is better than
cure." The percentage allowable will depend on the steel
and the purpose for which it is to be used. In a great deal
of ordinary steel it may reach .08 per cent, without serious
injury, but should always be kept as low as possible; in
other steel, for plates, etc., it frequently must be kept below
.03 per cent. The cold properties of steel are practically not
affected; the strength is increased slightly. Steel high in
sulphur will seldom get through the rolling mill. That the
sulphur is exceedingly injurious for very many purposes is
seen from the fact that red shortness will throw it out in
the mill from cracking, etc. For wood screws and gener-
ally where the product must be threaded, rather high sul-
phur is an advantage, say up to .1 per cent. This appears
to be due to the fact that the steel is less tenacious and does
not gall or tear as does tougher steel ; it also takes a better
polish. Manganese also helps in the latter process.

15. Pliospliorus. — In some respects phosphorus is the
most objectionable impurity in steel. Its most marked
effect is in producing a cold-short metal or one brittle at
ordinary temperatures. It does not affect the hot working
unless present in excessive amounts — .2 per cent., or higher.
It is objectionable here, as it gives a coarse grain to the
steel and lowers the point to which it can be safely heated.
Up to .12 or .13 per cent., phosphorus increases the strength
but lowers the ductility. The greatest objection is that
high-phosphorus steel is treacherous and is liable to break
under even small loads if suddenly applied. The behavior
of high-phosphorus steel is uncertain and whimsical through-
out, and for this reason its use is always perilous. The

14 MANUFACTURE OF STEEL § 36

ordinary limit in Bessemer steel is .1 per cent., but some
Bessemer is made as low as .075 per cent, phosphorus; by
the basic process it is usually made below .03 per cent.
Phosphorus is not known to be a benefit to steel under any
circumstances.

16. Silicon. — Silicon is generally absent in soft steels,
while in rail steel and castings it is present from .1 to .4 per
cent. In castings, it is added more to produce solidity than
for any effect on the physical properties. Soft and medium
steels, for plates, structural steel, etc., seldom contain over
.06 per cent, of silicoh and less than half of this usually.
There is some uncertainty and difference of opinion as to
the exact effect of silicon, but generally it does not affect
strength or toughness in amounts usually present. It
increases the stiffness, and is used in some heavy springs
requiring this feature. It also hardens the steel, and this
is commonly accepted as the beneficial effect in steel rails,
causing them to wear longer. Any considerable percentage
of silicon interferes with working at redness, welding, etc.,
and it is usually a cause of red shortness, although some high-
silicon steels forge well. All /the alloy steels having much
silicon must be worked at low heats.

17. Oxides or Oxygen. — These produce somewhat the
effect of sulphur, as cracking, and the effects of red short-
ness. Manganese removes them partially or nearly com-
pletely, depending on conditions in recarbonizing, and neu-
tralizes their effect in the steel. Their presence is greater
in soft steels or ones low in manganese, as they have the
oxidizable elements to seize in harder steels and thus be
removed as gases or solid compounds to go to the slag.

18. Copper, Nickel, and Aluminum. — Copper has
been supposed to produce red shortness in particular, but
later investigations disprove this, unless it is accompanied
by high sulphur, say .075 to .1 per cent. In amounts up
to .6 or .16 per cent, it has no effect on the cold properties,
unless adding slightly to its ductility, and only affects hot
working when sulphur or other red shortener is high.

§ 35 MANUFACTURE OF STEEL 15

Nickel steel finds its greatest use in armor plate, though
it has many other uses also, especially in high-grade for*
gings. Nickel has the property of giving a greater elastic
limit and ductility for the same tensile strength. Generally
from 3 to 3. 5 per cent, of nickel is present ; it is added in
the form of metallic nickel or ferronickel, in the melting
furnace.

Aluminum is seldom found even in traces in the amounts
added as a deoxidizer. Added in larger amounts, it increases
strength .somewhat and lowers ductility. Aluminum finds
no use except as a quieter in the proportion of from 2 to
6 ounces per ton of steel ; and this unites with the oxygen of
the bath and passes into the slag.

EXAMINATION OF THE FINISHED PRODUCTT

19. Chemical Examination. — After the steel is fin-
ished, it is subjected to examination to ascertain if the
desired qualities in chemical composition and physical prop-
erties have been reached. Throughout the manufacture of
steel the chemical laboratory plays an important part : First,
in the selection of proper materials; second, as a guide and
check in controlling operations; and finally, in the analysis
of the finished product. Methods of analysis are fully given
in Quantitative Analysis^ and represent the latest accepted
methods and those used in practical steel laboratories.
Many of the determinations must be completed in a very
short time to be of any value to the steel maker.

30. Microscopical Examination. — In addition to the
chemical examination, the finished steel is tested physically.
The microscope has been used largely in examining steel, and
has shed much light on its structure and constitution. It has
not been used, however, as a regular means of testing, mainly
owing to the time required to prepare sections for examina-
tion. It has been of practical value in detecting improper
heat treatment or in determining the proper heat treatment
for certain steels. It has been chiefly used with high-carbon

16 MANUFACTURE OF STEEL § 35

steels, and this seems to be its most favorable field, as the
carbon in such steels is most sensitive to heat treatment.
The microscope has revealed previously unexplained or
wrongly interpreted phenomena in the tempering of steel.

21. Physical Testing. — The steel is subject to various
mechanical tests for properties, such as bending, twisting,
quenching at redness and bending, tests of forgeability, etc.
Rails and axles are subject to drop tests ^ i. e., the fuU^sized
member (a section of rail) is supported near the ends on solid
blocks or foundations, and a weight, or /«/, dropped midway
between the supports. The height of drop and weight of
tup vary with the section of the member tested and the
specifications of the purchaser. Testing of this kind may
be regarded as qualitative^ so far as measuring the exact
force applied and expressing it in exact quantities. It is
not to be considered of less value or importance for this
reason, but that it is better adapted to show what the mate-
rial will do in service. All physical testing is, or should
be, made to approximate as closely as possible the actual
conditions under which the material is used.

23, Testing MacMne. — This may be defined as a
machine or apparatus for breaking samples of material and
measuring the stress required. The simplest conception is
to consider it as a weighing machine arranged to register
the force required to break or to produce certain effects in
the test specimen. It is used to pull test specimens from
plates, structural material, merchant shapes (rounds, squares,
etc.), of cast-steel test bars, etc. Fig. 1 shows one of the
standard types of machine of 100,000 pounds capacity.
They are made in all sizes up to 3,000,000 pounds capacity,
but above 200,000 pounds are mainly for experimental pur-
poses or special work; the 100,000-pound machine is the
size commonly used in testing laboratories. The machine
is driven by a direct-connected motor or from shafting.
Hydraulic testing machines were formerly much used, but
now hydraulic power is used only in the case of extremely
large machines.

§ 35 MANUFACTURE OF STEEL . It

23, The screw machine, Fig. 1, is the one used for testing
ordinary sections; in it the strain is applied to the piece
through vertical screws, one of which is shown at a. One

end of the test piece is held in the top, or fixed head d, of
cast steel, supported on cast-iron columns resting'on the

MANUFACTURE OP STEEL

M«

heavy iron base or weighing table, which, in turn, rests on
hardened-steel knife edges in a series of levers that transmit
the strain, as applied by the screws, to the weighing appa-
ratus ; the strain is registered by the poise on the beam c.
The lower end of the test piece is held in the movable or
pulling head rf, which is lowered or raised by the two
screws a^ reaching nearly to the fixed head, passing through
two brass nuts fastened in it; the screws pass down to the
base of the machine, where they are teyed to the main
gears by which they are revolved in either direction, raising
or lowering the pulling head as desired. Gears controlled
by the levers shown are provided for operating at several
different speeds. In both the fixed and pulling heads, holes
are cut with sloping sides in which wedges, or grips ^ fit for
holding the test piece.

In making a tensile test, the lower, or pulling, head is run
up to the proper height to adjust the specimen in the grips,
when the screws are reversed and the pulling head starts
down on the screws, stretching the piece until it breaks, the
upper end being firmly held by the grips of the fixed head.
The machine is principally used for making tensile or pull-
ing tests, but may also be used for compression or trans-
verse tests, when the grip lever and hanger on the pulling
head are removed, and the specimen placed on the weighing
table and the movable head run down on the specimen until
crushed or broken, the strain being registered on the beam
as in a tensile test.

24. Test Piece. — The standard test piece for most pur-
poses has a gauged length of 8 inches, in which the stretch

is measured. Fig. 2 {a)
shows the specimen for
plates and. structural
material for bridges;
ships, or buildings ;
and (*), the shape for
cylindrical bars. The
former is cut from the

§ 35 MANUFACTURE OF STEEL 19

finished plate, beam, etc., the edges being reduced as shown;
the two opposite sides are the rolled surfaces. In the case of
rounds, squares, rods, etc. , tests are made whenever possible
on full-sized sections as rolled and in a length of 8 inches.
For steel castings, forgings, and axles, the test specimen is
cut from the product and turned to a diameter of ^ inch by
from 2 to 4 inches gauged length. It was formerly quite
common to forge or roll a small ingot and make the physical
tests on this ; this is objectionable as heat treatment or work
received may be different and give varying results in tests;
test specimens are universally taken from the finished mate-
rial and tested in as near the natural condition, i. e., as
produced, as possible.

In test pieces that are machined, the opposite sides must
be parallel throughout the length of the test section, i. e.,
the length, or a little more, in which measurements are
made. Bars, rods, etc., tested in the shape they leave the
rolls, without any machining, usually vary but slightly in a
length of 8 inches and are calipered in several places and
the average taken. It is important that there should be
very little variation throughout the length, as it affects the
area on which the calculations are based. Measurements of
thickness, width, or diameter are made with a micrometer
gauge accurate to the one-thousandth of an inch, and from
these measurements the area of the cross-section is calcu-
lated— which in rectangular sections is merely multiplying
the two dimensions together; or in round sections, finding
the area of a circle with the diameter given. The elastic
limit and tensile strength, as shown on the beam of the
machine, are calculated from the area to pounds per square
inch, and always so reported.

26. Properties I>etennined In Testing:. — The prop-
erties usually determined in testing are (a) elastic limit,
(b) tensile strength, (c) elongation, (d) reduction of area.

The elastic limit is that point at which the metal under
strain takes the first appreciable set ; or the point at which
the steel under strain will not return to its original form

20 MANUFACTURE OF STEEL § 36

and dimensions when the strain is removed. This is by far
the most important property, as well as the one observed
first in testing. Steel strained beyond its elastic limit is
liable to give way under very light loads or much below its
original elastic limit; continued strains near the elastic
limit may produce the same result. It is determined in
testing by ** the drop of the beam '.' in all steel -works* labo-
ratories; automatic devices governed by electrical contact
are in use to a very limited extent. As the load on the test
piece increases, the poise is moved out along the beam to
just balance this; the instant the elastic limit is reached
there is a momentary and sudden elongation of the piece and
the load on the machine is released to such an extent that
the beam drops quickly in its surrounding guard. It remains
stationary a number of seconds, but the interval is decided
and lasts until the movement of the pulling head catches up
with the flow of metal in the test piece. In other words,
the metal of the test piece, at the point of elastic limit,
travels faster than the pulling head; hence, the drop of the
beam corresponds to the elastic limit. The weight shown
on the graduated beam is the elastic limit in pounds.

The term tensile strength is self-explanatory, and in deter-
mining it the stress is applied until the specimen parts. The
tensile strength is important in determining the fitness of the
steel for given purposes — but less so than the elastic limit.

Elongation is measured for most specimens in a length
of 8 inches (2 or 4 inches in castings and forgings). It is
determined by placing punch marks the proper distance
apart on the surface of the test piece before placing it in
the machine ; after breaking, the fractured ends are pushed
together and the increased distance the punch marks are
now apart over the original distance equals the elongation,
e. g., punch marks measure 10 inches apart after the frac-
ture of an 8-inch test piece, a stretch of 2 inches in 8 inches,
or an elongation of 25 per cent. ; it is measured to the
closest hundredth of an inch.

As the piece stretches, its cross-section is reduced and
the point where fracture occurs is drawn down, approaching

§ 35 MANUFACTURE OF STEEL 21

a conical point more or less. The area of this reduced sec-
tion, measured at the fracture, compared with the area of
the original section, is the reduction of area expressed in per
cent, of the original area. The elongation and reduction of
area are valuable expressions of the elasticity and ductility
of the steeL An increase of elastic limit and tensile strength
accompanies less elongation and reduction, or the harder
steels are stronger, but stretch and reduce less.

26. Effects of Work and Heat on Steel. — The physical
properties of steel are greatly affected by the amount of
work done upon it, and the temperature at which the work
is done. In general, the more work steel receives, or the
greater the reduction from a given section, the higher is
the elastic limit and tensile strength, with less stretch; but
the ductility (expressed in reduction of area) is not so much
affected unless there are great variations in heat at the same
time. In plates, with other conditions uniform, the thicker
the plate, the lower is the strength, and the less is the
stretch. Between a ^J-inch and a f-inch plate rolled from
the same steel, there may be a difference of 3,000 to
6,000 pounds per square inch in the tensile strength. This
difference may be further increased by working at a lower
heat, or lessened by rolling hotter. We have, then, increased
working adding to the strength, and, in fact, to the good
qualities of the steel, if it is done at the proper tempera-
ture. Cold working increases the strength, but at the
expense of ductility. Either extreme is objectionable, as
not developing the desired qualities in the steel.

27, The slabbing mill affords an advantage in making
slabs for plates over rolling large ingots directly on the plate
mill, as the latter method is apt to finish them too cold, with
its attendant disadvantages. With the slabbing mill, large
ingots are rolled into slabs adapted to the size of plate
to be made. This allows different sized plates to be finished
nearer the same temperature, gaining the advantages of
increased rolling and avoiding the evils of cold rolling.

82 MANUFACTURE OF STEEL § 35

The latter is not always a disadvantage, but is in the class
of material considered. In certain finishing mills, sheets or
other products are regularly cold rolled where a sacrifice of
ductility is of less consequence than the strength gained ; it
also produces a denser, stiffer, and harder product. Rails
have lately been cooled somewhat before being put through
the finishing passes, the result being a denser and harder
metal, particularly in the head, which is expected 'and
claimed to increase their wearing power. This is especially
shown in the very heavy rails where there is a larger mass of
metal.

!38« Belation of Chemical Composition to Streng^tli.

Much work has been done by various investigators to estab-
lish the relation between the chemical composition and the
strength of stee^ and various formulas for calculating the
strength from the composition have been proposed. Mr. W.
R. Webster has conducted the most exhaustive experiments
in this direction, and his results in many cases quite closely
approach those obtained from the testing machine. With
all conditions uniform — the same steel, equal size ingots or
slabs, heated to a like temperature, and the amount of
reduction in rolling, etc. — the chemical analysis will give
the strength very closely. But, owing to variations in mill
practice (principally finishing temperature and different
amounts of work), some of which cannot always be kept
within the close limits desirable, the estimation of strength
from analysis may be said to be only an approximation.
However, this approaches so nearly the results of tests that
it is of great value as a preliminary estimation of the ulti-
mate strength.

Table V is based on Webster's results, and from it the
approximate ultimate strength can be found. It is worked
out for the elements, carbon (up to .25 per cent.), phosphorus,
manganese, and sulphur (in amounts usually present), with
corrections for different widths and thicknesses. A brief
study of the table will show the manner of applying it.
The example given herewith illustrates it:

§35

MANUFACTURE OP STEEL

^
^
I

C :

•B 8 1

5 ^ 4)

•

(2*0,

C3 en

O 0) "M

•13 JS C

as '3

•

C "^ u

O CO 4)

e«

S.S ^

ij^: «

•

•^^
M

Ultima
Strengt

Pound

PL.

••

o . en

•V4 ' * -

5 .t; •o

J.g §

e«

VO •

e«

•*

»<

»•

».

r«

«^

•<»

•k

»•

»>

WJ £

1 £

•2 u

? 4>

C/D PL,

i s c

J3 0 ^

•

V> W

S) "

'i-

1 <^

•

s «=

•§ o

'i-

<« u

•

o £

•H

e«

r««.

3 H

MANUFACTURE OF STEEL

8 35

V JO noilanpsH

i s R

» s,

1 ! 1

1 i

■«»ni|3!qj,

1 M

iSl|3QI

M s

r?^^

8 8 S

S, S 8

i
1

8. '^ ■?

s a ^

E
1

^^UI » J sabs

i S i

144

■punod

1 1 1

s s s
Hi

1 i 1

K R S

1 1 1

1 5 1

1
1

■™v

^ ^ i

III

■SS3U1|3!'1J.

■« S S

-mpsajg

sis

ff* t

I
1
1
1

-^1

sis

•■II

s 8 r

i s a

^1 £

" r ?

? « J

B9i

o.«,<unK

, ~ ^

" "■ a

§ 35 MANUFACTURE OF STEEL 25

Example. — A given specimen analyzes: carbon .21 percent, phos-
phorus .035 per cent., sulphur .082 per cent, manganese .36 per cent.
The plate is 80 inches wide and ^^ inch thick. Finding the carbon in
the upper horizontal line of Table V, and going down this column until
opposite per cent, phosphorus (left-hand vertical column), we find
56,800 pounds per square inch as the strength for .21 per cent carbon
and .035 per cent, phosphorus. The addition for .032 per cent, sulphur
is 1,600 pounds; for .36 per cent, manganese, 7,960 pounds; an 80-inch
plate 1^ inch thick calls for a deduction of 250 pounds. We now have
[56.800 (C-hP)-h 1,600 5+7,960 Afn] - 260, or 66,110 pounds, as the
ultimate strength per. square inch. In using the table, the differences
in strength due to the varying temperature and the rolling must not
be forgotten, and as these cannot be allowed for, the results from
calculation are always liable to differ from those obtained by pulling
specimens in the testing machine.

29. Results In Physical Testing. — Table VI gives the
composition, together with the results obtained in testing a
number of steels of varying carbon percentages.

Table VII shows a record of physical tests, with the
measurements, results, etc.

Numbers 9 and 10 of the table are steel castings, pulled
in a length of 2 inches, and turned to the diameter shown.

RECENT PROGRESS IN STEEIi MAKING

SPECIAL. METHODS

30. General Remarks. — While many radical improve-
ments and modifications have taken place in the Bessemer
and open-hearth processes, no fundamental principle invol-
ving an entire change of method and apparatus has been
developed since their introduction. The most distinct as
well as the most promising attainment is t/ie Talbot continu-
ous open-hearth process^ developed and patented in 1899 by
Benjamin Talbot, of Pencoyd, Pennsylvania.

31. Talbot Continuous Open-Hearth Process. — This
consists in maintaining a constant reservoir of metal, part of
which is withdrawn when completely refined, and the same

26 MANUFACTURE OF STEEL § 36

amount of liquid pig metal added from the cupola or the
blast furnace — the latter preferably, as it avoids the cost
and labor of remelting. The process is carried on in a
slightly modified rolling or tilting open-hearth furnace.
The original one is of 75 tons capacity, 20 tons of steel being
poured off for each heat and then 20 tons of melted pig
iron added to the 55 tons of steel in the furnace ; scale, cin-
der, ore, and limestone, to form a slag, are added between
each withdrawal and addition. Furnaces of 150 and 200 tons
capacity are under construction in America and England;
one being in as successful operation in England as the orig-
inal smaller one. It was at first feared that the bottom
could not be kept in proper condition, owing to the large body
of metal and the fact that the bottom could not be reached
until the furnace was emptied at the end of each week.
This has been proved erroneous, and what some metallur-
gists believed, that it was not the metal but the slag that
injured the basic bottom, was proved. The bottom does
not suffer except by scorification at the slag line, as in the
ordinary practice. It is repaired as usual. In fact, the
scorification is somewhat less, as part of the slag is decanted
before the steel is poured out. It is essentially a pig, or
pig-and-oxide, process, as little or no scrap is used, and this
point makes it advantageous or otherwise, according as pig
iron or scrap is the cheaper and more abundant stock.

32. About the only difference in the furnace from the
usual tilting one is that it is made with a slag spout at the
back and to tip both ways, so that the slag may be decanted
off from the side opposite that where the metal is tapped,
into a slag ladle or car beneath. The construction is neces-
sarily stronger to correspond to the increased weight of
metal, etc. The initial or filling heat (at the beginning of
each week's run) is prepared in the usual way from scrap
and pig metal — the latter preferably added molten on top of
the scrap and limestone — worked down to steel, feeding ore
if necessary, as in common practice; when thoroughly
refined and in proper condition to tap, about 20 tons (from

§ 35 MANUFACTURE OF STEEL 27

one-third to one-fourth of the bath) is poured off, recar-
bonized in the ladle, and cast into ingots. Before adding
more metal, oxide of iron in rather fine condition is added
to the slag, and as soon as melted a part of the 20 tons of
metal is added ; when the violent reaction from this ceases
somewhat, a second addition of oxide of iron is made,
together with limestone; then more metal, and a third addi-
tion of oxide with stone, and finally the third and last metal
addition. From this point the heat is handled as in ordi-
nary practice until the entire contents of the furnace — not
merely a part of it — is reduced to good steel minus the
recarbonizer. Twenty tons are again poured off, iron oxide,
limestone, and metal added as above; this round of opera-
tions being kept up during the week and the furnace emptied
on Saturday.

The details of operation vary somewhat. As the entire
amount of oxide and stone may be added before the metal,
and this added in two or three pourings, as the violence of
the reaction permits, the time between additions may vary,
etc. As a rule, the first two metal additions are about equal,
and from 80 to 90 per cent, of the total pig metal ; the third
addition (about 15 per cent.) is usually added from 10 to
20 minutes before tapping, depending on the amount of
carbon wanted in the steel and the rapidity of the reactions ;
an interval of from 2 to 2^ hours generally elapses between
the second and third metal additions, although this is not
an essential and all the metal may be transferred to the
bath within an hour. Forty-two heats per week have been
made; between taps there is an interval of about 4 hours
against about 8 or 10 hours in the ordinary practice.

Approximately the same tonnage per week has been made
as in the largest type (50 tons) of ordinary furnace, when the
entire bath is tapped, and using pig and scrap with some ore.
A much larger output is counted on with the 150-ton to
200-ton furnaces; this and the fact that the process can be
worked with pig and oxides or ore are its chief advantages;
minor, but important, ones are that the yield is greater and
that a wider range of pig iron can be used (higher in silicon

MANUFACTURE OF STEEL

§35

and phosphorus). The last is due to the fonditions under
which oxidation occurs and that the excess of slag can be
removed as formed. The increased yield of steel comes
from the iron reduced from the oxides, entering the bath.
The usual basic conditions of a minimum of SiO^ in the
stock must be observed, but this is of less importance than
when the slag cannot be removed.

The iron oxides used are roll scale, mill cinder (basic low
SiO^y and iron ore; ordinary limestone, if low SiO^. The
composition of the oxides is shown in Table VIII.

TABIiE Till

Material

Iron.
Per Cent

Silica.
Per Cent.

Phosphorus.
Per Cent.

Roll scale

70 to 74
64 to 68
60 to 65

•5 to 3
3.0 to 8
3.0 to 8

.05 to .2
.05 to .5
,05 to 1.0

Basic mill cinder. . .
Iron ore

33. The process is based on the powerful oxidizing
action of a slag rich in iron oxides ; or stated the other way,
the prompt reducibility of slags, rich in iron oxides, by the
metalloids in the bath. In the usual process, much of the
oxidation is done by the oxygen of the air, whereas here
the slag (or the oxygen in it) performs this work. The
reactions are as follows :

FeO+C
^ZFeO + Si
6FeO + 2P

CO+Fe
SiO, + 2Fe
Pfi^ + hFe

This is one of the most active and rapid reactions met
with in iron and steel metallurgy; in fact, it is only paral-
leled, if equaled, by the rapid purification in the Bessemer
converter. This is shown by the following example from
the records of operation.

Example. — Into a bath of 104,000 pounds of metal ready to pour
and covered with a slag containing 22.4 per cent, of metallic iron.

§ 35 MANUFACTURE OP STEEL 29

9,300 pounds of liquid pig metal was poured, taking 4 minutes;
2 minutes afterwards the metal was hot enough to tap and was thor-
oughly purified — or in 6 minutes 9,300 pounds of pig metal was
converted into steel. At the same time the iron in the slag was
reduced to 12.4 per cent., nearly one-half the oxygen in the FeOyFe<xO%
being given up to oxidize the metalloids. During the reaction the
furnace doors were opened and the movable ports pushed back to allow
the escape of the gases formed by the reaction. All the gas was also
shut off the furnace until the action quieted down. This reaction had
all the characteristics of the Bessemer blow. While carbon was being
burned, a large volume of carbon monoxide was given off,

FeO + C = CO + />,

which ignited and burned with an intense heat, a part being absorbed
by the bath and a part raising the temperature of the regenerators.

The process yields about 105 per cent.; i. e., for every
100 tons of pig metal charged, about 105 tons of steel is
poured out, the increase coming from the iron reduced by
the metalloids from the oxides in the slag.

34. Monell Process. — This process was developed a
little later than the Talbot and may be briefly and to some
extent described' as the latter worked in the ordinary station-
ary furnace, all the metal being tapped out at once. It was
worked out at the Carnegie Steel Company's works and is
used to a considerable extent by them. It involves no new
principle; in fact, the same method was tried in the early
history of the basic open-hearth process, but Mr. Monell has
achieved much greater success than ever before reached by
the method. As worked at the Homestead plant, limestone
and iron oxides (ore, scale, or low-silica cinder) are charged
on a basic hearth, heated to partial fusion, and liquid pig
iron poured in, when the action becomes violent and the
metalloids are rapidly oxidized. This action is the same as
in the Talbot process — except much less intense — and is due
to the slag, containing the excessive amount of iron oxides.
The slag may be tapped off through a tap hole placed above
the level of the metal, but this is not within the easy control
obtained by decanting from a tilting furnace, and is one of
the objections to the process. Another is, the slag, rich in

30 MANUFACTURE OF STEEL § 35

oxides, corrodes the bottom, if it comes in contact with it,
and this cannot be entirely avoided. About the same or a
slightly increased output per week is obtained over the same
furnace using pig and scrap. The yield from metal charged
is less than in the Talbot, about 102 to 103 per cent. The
same stock is available as in the ordinary open-hearth proc-
ess or the Talbot, but the latter process allows a wider
range of silicon and phosphorus to be used.

35. Bertrand-Thlel Process. — In this process two open-
hearth furnaces are operated as a unit and the metal trans-
ferred from the first, or melting, furnace, called the primary^
or refiner^ into a secondary^ ox finisher ^ furnace. It is the
invention of Messrs. Bertrand and Thiel, Kladno, Austria,
and has been in successful operation there since 1894. While
only one other plant (in England) has been constructed to
operate on this system, several others are shortly to be
built in Europe, and probably one or two in America. It
may be worked on either the acid or basic hearth, but so
far has only been worked on the latter, and is not likely to
be used for acid practice.

One of the chief advantages of the process is its flexibility,
as it may be worked exclusively as a pig-and-ore process or
pig-and-scrap with equal advantage and in whatever propor-
tions available. As in the former process, the refiner is
charged with liquid pig iron and enough lime or limestone to
furnish a basic slag and ore to oxidize part of the metalloids;
the amount of the latter (mainly silicon and phosphorus) in
the metal determines the amount of lime and ore to be used.
In this furnace all of the silicon is oxidized, approximately
90 per cent, of the phosphorus and manganese, and about
40 per cent, of the carbon. As stated elsewhere, both silicon
and phosphorus are oxidized at comparatively low tempera-
tures; this accounts for the removal of these elements in
the first furnace. The metal is then transferred to the
finisher, into which has previously been charged about half
the quantity of lime or limestone, and ore used in the
refiner, so as to be heated nearly to the fusing point. The

§ 35 MANUFACTURE OF STEEL 31

hot metal, with from 2 to 2^ per cent, of carbon, no silicon,
little manganese, and a small percentage of phosphorus,
coming in contact with the highly oxidizing slag has the
carbon and remaining phosphorus quickly removed. It will
be seen that the oxidation of the metalloids in the Talbot,
Monell, and Bertrand-Thiel processes depends on the same
principle — the oxidizing power of a basic slag rich in oxides
of iron — though applied somewhat differently in each case.
In coming from the first to the second, or finishing, fur-
nace, the slag is skimmed off and very little allowed to enter
the latter. As originally worked, the refiner furnace stood
on a higher level than the finisher — both being stationary —
and the metal run down a trough. This is not an essential
feature of the process, and either stationary or tilting fur-
naces on the same level may be used, the metal being trans-
ferred from one to the other by ladle and crane. The tilting
furnace, on the same or a higher level, offers the advantage
that the slag may be conveniently handled by decantation.
In case scrap is used, a small amount is charged into the
refiner, and the greater part into the finisher, with the stone
and ore, and allowed to heat and oxidize somewhat before the
refined metal is added. This oxidation of the scrap is not a
loss, as it takes the place of some ore, and the carbon and phos-
phorus reduce it to metallic iron, which is added to the bath.

36. About the same yield as in the Monell practice is
obtained — 102 or 103 per cent, of the metal charged. An
output of 45 heats per week from the two furnaces has been
obtained, a greater number than from two furnaces using
similar stock worked on the usual system ; the tonnage has
been much less than from two large furnaces, as only small
ones (20 tons) have been used so far; but there is every
reason to believe that nearly as many heats can be made by
using large furnaces — when the tonnage will be greater than
that obtained from two of equal capacity — and the operation
finished in one furnace. The charge is in the first furnace
2 or 3 hours, and in the second from 2 to 2^. This can be
adjusted, however, by the point to which the refining is

32 MANUFACTURE OF STEEL § 35

carried in the former. In the finisher, the heat is boiled
down as in ordinary practice, using ore, if necessary ; when
the proper temperature is reached, the heat is tapped into
the ladle, recarbonized as usual, or this may be done pre-
viously in the furnace and cast into ingots.

37. Duplex Process. — This process takes its name from
the fact that both the Bessemer and the open-hearth appa-
ratus are used to produce the steel. It was originally and
is still used at Witkowitz, Austria, but has been installed in
other parts of Europe, particularly in the Middlesbrough
District, England, and experimentally in America, but its
application has been limited. Its field is in converting pig
iron too high in phosphorus for the acid process and not
high enough for basic Bessemer practice, or too high in
silicon for the latter or the basic open-hearth. In using it,
pig metal is melted in the cupola, or taken direct from the
blast furnace, and is blown in an acid Bessemer converter to
remove practically all of the silicon, part of the manganese,
and a little of the carbon, the phosphorus not being affected.
The desiliconized metal is poured into a ladle and trans-
ferred to a basic open-hearth furnace, where dephosphor-
ization and decarbonization take place, as usual. The
process was advanced by many metallurgists as the proper
one to convert into steel the rather high-silicon and phos-
phorous pig irons of the South (Alabama, etc.), but better
blast-furnace practice, giving low silicon, has made the metal
suitable for the basic open-hearth process, which is now used.

TREATMENT OF THE INGOT

MTL,I^, ETC.

38. General Remarks. — As this is a distinct subject
and more especially a mechanical one, merely an outline of
it can be given. Only the mills that receive the ingots from
the steel-making departments, or plants, and work them

§ 35 MANUFACTURE OF STEEL 33

down for the great number and variety of finishing mills, or
a few finished lines, as rails and plates, will be mentioned.
The ingots, after having been poured into iron molds
placed in a casting pit or on cars, are taken to the
heating furnaces to be reheated for rolling. Casting on
cars is the later practice, and is followed at most up-to-
date works, as it saves the labor and expense of pit casting,
besides being a more expeditious method of handling the
steel. A pit is required where ingots are cast in groups,
the molds being filled from the bottom through a center
runner with connections to each mold; such bottom-cast
ingots are necessary where they are rolled direct from the
ingot into plates, as the ordinary top-cast ingots will not
give as good a surface when rolled at one operation into
plates. When cast on cars, these are shifted to the stripper^
an hydraulic or electric mechanism for removing the molds
from the ingots. This is done as soon* as the ingots solidify,
so as to get them into the furnace promptly, and thus to
require the least reheating to bring them to the proper
rolling temperature.

39, Relieatlni? Furnaces. — Reheating furnaces are
either horizontal or vertical, the latter being used almost
universally for heavy ingots, and the former for lighter
sections, slabs, blooms, etc., for rerolling. Both types are
equipped with Siemens regenerators, the necessary flues,
reversing and controlling valves, etc., the same as is the
open-hearth furnace. The vertical furnace is commonly
called a soaking pit ^ but more correctly a pit furnace. It is
identical in construction with the crucible melting furnace.
Fig. 8 (a) and (^), Part 2. It is divided into holes in the
same way for 4 or 6 ingots placed on end — corresponding to
the crucibles in the melting furnace. A furnace is usually
built with from 4 to 10 holes and is designated by the num-
ber of holes — as an 8-hole pit furnace. In all late construc-
tion each hole has its own separate air and gas flues
controlled by independent valves (or two holes may be
connected). This permits adjusting the temperature in

34 MANUFACTURE OF STEEL § 35

different sections of the furnace, as required by the steel.
There is only the one set of air and gas reversing valves for
the furnace, so that the currents of gas and air are reversed
at the same time for all the holes.

40. Mills. — In general from their construction, mills are
designated as 2-high^ S-high^ reversing^ non-reversing^ or
universal. The first two indicate the number of rolls in the
same vertical plane ; the second two as to whether the rolls
are driven in the same direction all the time or reversed at
intervals; a universal mill has the regular horizontal rolls
and, in addition, vertical ones, so that the piece is rolled on
its four sides. Two-high mills (of this class) are always
reversing; and 3-high(ofany class), non-reversing; univer-
sal mills may be 2-high reversing or 3-high non-reversing.
From the purpose for which they are used, mills are known as
blooming mills y plate mills ^ slabbing mills; in Great Britain,
the latter are known as cogging mills. It will be remem-
bered this*does not touch the field of finishing mills, but only
mills rolling ingots into finished or intermediate products.

41. Bloomliig Mill. — This mill breaks the ingot down
to blooms, billets, or slabs.

A bloom is a section of the reduced ingot to be finished
on a succeeding mill ; they are generally square, but may
have the shape roughed out into which they are to be
finished. Their size and weight vary with the purpose for
which they are used. They are usually for rails, structural
shapes, or forgings.

A billet is a smaller section of the ingot — or a small
bloom. They are used for merchant shapes — rounds,
squares, etc. — and rods for wire, bolts, rivets, etc. . Their
size varies, but the standard section is 4 inches square
and of different weights.

A slab is a fiat shape of varying dimensions that is to be
rolled into plates or sheets.

Blooming mills are built 2-high reversing or 3-high non-
reversing, generally the former ; these are driven by power-
ful reversing engines through pinions, the ends of these

! 36 MANUFACTURE OF STEEL

36 MANUFACTURE OF STEEL § 35

connected with the ends of the rolls. The rolls are carried
in heavy iron or cast-steel housings, the bottom one on
stationary bearings, the top one balanced by counterweights
underneath, which keep it against vertical screws in each
housing by which it is raised or lowered; the screws are
operated by an hydraulic cylinder applied through a rack
and pinion. In 3-high mills the middle and upper rolls are
usually both movable vertically, and the bottom fixed.
Fig. 3 shows a 2-high, reversing, blooming mill dy together
with the roll tables b. The mill engine is not shown; the
small one a at the end operates the tables, which are always
driven separately from the mill proper. Hydraulic manipu-
lators from beneath turn the ingot on the tables; the opera-
tion of the mill and engines is controlled from the raised
platform, or pulpit c^ on the left. In all 3-high mills, the
roll tables are made to raise and lower by hydraulic
mechanism, so as to bring the piece opposite the passes
between the top and middle or bottom and middle rolls,
as required. In rolling rails, the ingots are bloomed down
to the size rail bloom required for one, two, or three rails;
after a short reheating, these go to the rail trains, which are
3-high non-reversing mills, the first, or roughing, train forms
the rail, and the second, or finishing, train (or rolls) com-
pletes the operation. The trains are generally arranged
tandem, so that the passage from one mill to the other is
continuous. Structural shapes are rolled in the same gen-
eral way, with various modifications in the type and arrange-
ment of mills.

42« Plate Mill. — These are mostly 3-high non-reversing
mills, but are made 2-high reversing for lighter work;
most universal plate mills are also so built. Fig. 4 (^i)
shows the mill proper and connections from the pinions, the
bottom or middle one of the latter in this case being con-
nected to the engine. Fig. 4 {b) shows a side elevation of
the tables, with the mechanism for raising and lowering
these, and the end of the rolls in the housing. In plate
mills, the middle roll is not connected to a pinion, but is

§ 35 MANUFACTURE OF STEEL 37

driven by the friction of the top or the bottom roll. Three-
high plate mills are built in all sizes up to 132 inches for
finished length of rolls, permitting plates about 10 J^ feet
wide to be rolled.

43. Universal Mill. — As already explained, these have
vertical and horizontal rolls, so that the piece is rolled on all
four sides at the same time. They are built for plate mills
and slabbing mills.

44. Universal Plate Mill. — This mill may be either
2-high or 3-high, and is used especially for long and narrow
plates, such as bridge plates, pipe skelp, etc. ; rolls are
adjustable in both directions for various-sized plates. The
general width of universal plates is from 6 to 48 inches.

45. Slabbing: Mill. — A slabbing mill is a universal mill
for rolling down heavy ingots into slabs for plate mills, thus
relieving the latter of the work of breaking down ingots for
large plates, increasing the tonnage, and reducing the scrap
made. The advantage of the slabbing mill is to increase
the output of the plate mill and also permit the making of
top-cast ingots for plates, and also larger ones, at the steel
plant. The extra work the steel receives is also an advan-
tage. Fig. 5 (a) and (d) is a view of a 2-high universal mill
showing vertical and horizontal rolls; (a) is a front eleva-
tion showing rolls, pinions, and connections, while (d) is a
side elevation showing the table and the rolls in the hous-
ing. The construction of such mills for either a plate or
slabbing mill is practically the same, except that the latter
is built much stronger and heavier.

REFRACTORY MATERIAIiS

46. General Remarks. — The success of steel-making
operations, among many other factors, depends on the
ability of the apparatus to withstand the heat conditions,
and not a little of it is due to the refractory materials.
The metallurgist needs to be no less familiar with their

MANUFACTURE OF STEEL

§36

properties than with the reactions and manipulations of the
processes themselves. Though there is quite a range of
materials used for lining for furnaces, vessels, ladles, etc.,
they may, according to their chemical nature, be divided as
follows :

Acid materials. . . i o-i- • ^ • t

( Sihcious materials

Refractories

Basic materials. . .

Neutral materials

Magnesite

Dolomite

Limestone.

Chromite

Carbon

Bauxite

These materials may be used either in mass, shaped to the
purpose as applied, or as brick ; both forms are essential and
extensively used.

ACIB REFRACTOBTES

47. Clays. — The most important of the first class is
clay. Clay is a hydrated silicate of alumina, always con-
taining varying amounts of free silica, oxide of iron, lime,
magnesia, generally alkalies, and frequently titanium. A
true clay has the composition SiO^y 46.4 per cent.; Al^O^^
39.7 per cent.; H^O^ 13.9 percent. ; this corresponds to the
formula Al^Sifi^.'lHfi^, or Alfi^.^SiO^.lHfi. Pure kao-
lin (china clay) represents this composition, but owing to
its cost, it finds no use in steel refractories, except to a
small extent in the manufacture of crucibles. A few clays
approach this composition clogely, and it is the ideal one, as
it represents the least fusibility when free from injurious
impurities, the SiO^ and Alfi^ being present in almost
exactly the proportions to give greatest infusibility attain-
able in a clay. Clays free enough from impurities to stand
a high fire test are used for making brick and for ball stuff,
and are known as fireclays. It is only these with which we
are concerned.

§ 35 MANUFACTURE OF STEEL 89

Clay results from the atmospheric decomposition of vari-
ous rocks, mainly feldspar, orthoclase Kfi^Alfi^y^SiO^\
this is broken down by the action of air, moisture, and CO^
to a soft mass that absorbs water, the potash and part of
the silica is leached out and carried away in solution or
mechanically. Pure kaolin would result from the decompo-
sition of the above, but the silicates generally contain other
minerals with />, CaO^ MgO^ etc. ; so we find these in the
clay. They belong to all geological periods. The more
recent are softer and more plastic, having all their com-
bined water, while older ones appear solid and dry from
having lost their hygroscopic water; still older ones, that
have possibly been subject to great heat and pressure, have
also lost their combined water, and with it the property of
becoming plastic by the addition of water.

48, Most clays used for refractory purposes occur in the
coal measures, frequently under or between the coal veins.
This does not mean as good refractory clays are not found
elsewhere, but is a matter of geographical location, as the
situation of iron and steel industries is generally determined
by their closeness to fuel; hence, suitable clays that are
nearest are used. Nearly all fireclays contain more silica
than the formula for a pure clay (as represented by kaolin)
calls for; they might be called acid or silica clays; but such
a distinction is not observed. This excess of silica is not
especially objectionable in most cases, although it lowers the
refractoriness, but must not be present in sufficient quantity
to lessen the plasticity very much. The two chief points in
a fireclay are refractoriness and plasticity. The necessity of
the first is apparent and requires no comment. Plasticity is
essential, that it may be molded into the desired form and
retain this while drying and burning — ^as brick or ball stuff
for lining, patching, etc. The plasticity of clay is due to
the combined water contained. If air dried, or dried at a
temperature of 100° C. (212° P.), it appears hard and thor-
oughly dry, but still contains its water of combination. On
beating to redness, this is lost, and it becomes a hard mass —

MANUFACTURE OF STEEL

§35

U, "M

00 VO

m ^ M

M On

M W

On VO JS
to lO ^^

1> V O 4^ 4)
'O "TD 'O 'O 'O

u->

•

■

•

u->

c«

u->

•

Q. 4)

On 00

00
O

On
O

00
VO

O O

ro 00

■ •

On 00
00 M

O On
u-> vO

00 u->0 r*tO'4*0 MOO t^w

.♦* CL 4)

N lO

rO VO

ro O
rO VO

O 00 VO
00 fO u->

lO O

U-) 00

rO VO
On 00

o J^

t^
ro

w
ro

O U-) to M O fO 00
IT) lO VO VO VO vO t^

M t^

M 00

to VO
C« VO

ro to 00

On On t^ *0

4J
O
U

o
(/}

u
G

0)
N
4;

a
H

<4-»

c
4;

4)
N
4)
G
4)

rt rt

G
>

C
>

'-5 .— ►^ »>

4;
bo

C
>

pL,

O
>
O

c
4;

c/3 pu!

C
G
4;

Plh

o
>

G
4;

G
C
4)

C
G
4;

ctf

bo .3

Ph Ph :=:

G -^
.G tn

^ 4J

G 4>

4) :i:
•^ 4;

; - ^

G
G
O

"* G
O

'3
PL4 P^

CTS
CA

G
C

4>
PL|

o
a

CTS
CA

^ 4>

>; ^

.h O

oidra^S JO

M N

to VO

00 OS

c«

§ 35 MANUFACTURE OF STEEL 41

biscuit — which is very porous and will absorb considerable
water, but cannot be made plastic again. When air dried,
or dried at 100** C, plasticity is lost, but is restored by mix-
ing with water.

Table IX shows the analyses of typical and well-known
clays.

The most objectionable elements in fireclays are alkalies,
iron oxide, lime, and magnesia, and their bad effect is in
about this order for the amounts usually present. The
alkalies (K'^O and Na^O or salts) are the worst and act by
forming readily fusible alkaline silicates. Traces are prob-
ably present in all fireclays; the amount is usually under
1 per cent. Oxide of iron can do no good, but its presence
in moderate amounts does not seem to greatly lower the
fusibility. If present in the clay as FeO^ it is more harm-
ful than if as Fe^O^^ as the latter does not combine with
SiO^y while the former does as ferrous silicate; if brick are
burned or used in a highly reducing atmosphere, the Fefi^
may be reduced and the silicate formed. Iron gives a reddish
or brown color to the brick or clay on burning. . Lime and
magnesia are usually present in varying percentages ; they
are less objectionable than iron or alkalies, but add somewhat
to the fusibility. Titanic acid TiO^ is present in some excel-
lent clays; its effect is not well understood, but under cer-
tain conditions it adds to the refractoriness.

49. SllldoTis Materials. — In these silica predominates,
being about 90 to 99 per cent. , as a rule. Silica rock fur-
nishes most of the material under the heads of silica brick,
ganister, and silica sand. For the manufacture of silica
brick the rock contains 98 or 99 percent. SiO^^ the remainder
being Fefi^^ -^A^a» CaO, MgO^ and sometimes combined
water. For brickmaking, purity of the rock is not the only
requisite; m fact, some rocks of nearly pure SiO^ are unfit
for the purpose — mainly owing to the brick expanding
irregularly and excessively when exposed to a temperature
above that of burning. All silica brick at high tempera-
tures expand about \ inch to the foot, but it must be

MANUFACTURE OF STEEL

§35

uniform and take place gradually. As the ground rock has
practically no plasticity, binding material must be added to
fit the particles together in burning ; from 1 to 2 per cent,
of lime or a refractory clay is generally used. The bricks
are carefully dried and then fired at a high heat for about a
week and allowed to cool in the kiln for about the same time.
The temperature of kilns burning high-grade refractory brick
of any kind is commonly controlled by the use of Seger cones,

Ganister is a silicious rock generally containing some-/4/,(?,
and combined water. Its principal use is for converter linings
in the acid Bessemer process, and mixing with clay for ladle
linings, etc. It may be regarded as silica with enough clay
material to bind it, giving a strong material when burned.
Quartz has been used for the same purposes.

Silica sand is nearly pure SiO^^ used mainly for bottoms
or hearths in the acid open-hearth process; also for bottoms
of heating furnaces.

Table X shows the composition of silicious materials.

TABIiE X

3
4
5
6

7
8

Silica rock

99.11

Silica rock

96.90

Ganister

98.72

Ganister

84.60

Silica brick

96.52

Silica brick. ...

94.82

Silica sand

98.30

Ouartz

94.20

AUO^

Fe^O^

CaO

MgO

Per

Cent

Cent.

.21

.64

.04

2.00

.50

.06

.59

.20

.16

trace

11.80

.68

trace

1.40

.60

1.48

.06

.86

.50

3.82

trace

1.02

.58

trace

2.10

1.60

.40

.60

Com-
bined
Water.

Percent.

.65

.24
2.80

.25
.90

BASIC REFRACTORTBS

50. Magrnesia is the most important of this class. It is
used principally for the hearths of basic open-hearth furnaces,
lining for basic Bessemer converters, and for making brick

§ 35 MANUFACTURE OF STEEL 43

for similar uses. It occurs naturally as the carbonate (mag-
nesite), Greece and Austria furnishing practically the supply
for the world. That from Greece is the purer, but is not as
well adapted for basic linings as the Austrian, but makes a
superior brick. It is always used as the oxide and requires
an extreme heat to drive off the last of the CO^, One objec-
tion to the Grecian is its freedom from other bases that lower
the fusing point, it being too refractory for many purposes.
The natural Austrian (the best known is that of Karl Spaeter)
seems to have the impurities blended in about the correct
proportions for the best results and is sufficiently refractory
to set well and give a hard bottom that does not wear readily.

Dolomite, or magrnesian limestone, is scarcely less
important than magnesite. They are used to a large extent
interchangeably. Formerly, basic Bessemer converters and
basic open-hearth furnaces were universally lined with dolo-
mite, but magnesite has now taken its place, being superior
to it. Dolomite is used where a somewhat less refractory
material is wanted, usually for patching vessels and hearths;
here it has an advantage over magnesite, as it sets quicker,
shortening delays for repairing, and lasts nearly as well. It
is burned in cupolas or kilns to expel the C(?„ and absorbs
moisture if exposed to the air for a considerable time; but
under ordinary conditions may be kept 1 to 2 weeks.

liimestone is the cheapest and most abundant basic mate-
rial and is extremely refractory, never having been fused or
even softened, but it cannot be used, as it is next to impos-
sible to get it to bind, and it cannot be kept, as it absorbs
water so rapidly ; hearths or linings made of it, if left stand-
ing, soon disintegrate from the slaking of the lime. It is
the principal flux in basic-steel making.

NEUTRAL. REFRACTORTES

61, At one time neutral' substances were looked to as
linings, so that either an acid or basic process could be
worked in the same apparatus. It has not been success-
fully accomplished, nor is it likely to be. Later, it was

MANUFACTURE OF STEEL

§35

considered essential to have a neutral band between the
basic hearth and the silica side walls. This was found to
be unnecessary, and the silica brick, in the basic open-hearth
furnace, are now laid directly on the magnesite, the latter
being carried above the slag line.

Chromlte is the most valuable of the class of neutral
materials, and is used mostly for patching thin walls of basic
furnaces where silica would be likely to get to the bottom;
chrome bricks are used to a certain extent as a layer under
the magnesite brick in the basic hearth ; and to some extent
in reheating furnaces where the fluid cinder — ferrous sili-
cate, principally — cuts other brick. A serious objection to
using much of it in melting furnaces is that some of the
chromium is reduced and absorbed by the metal, producing
hardness, especially hard spots. Chromite is a double oxide
of iron and chromium, Cr^O^^FeO^ with varying amounts of
silica and other bases. The supply comes principally from
Canada and Turkey.

TABIiE XI

0-,(9,

FeO

Al^O^

S/O^

CaO

M^O

TiO^

Material

Per

Cent

Cent.

Chromite

5123

36.63

3.17

1.87

5.10

3.79

Chromite

62.20

28.10

2.60

3.07

1. 10

Bauxite

1. 00

90.00

2.00

1.75

trace

5.00

Material

Graphite
Graphite
Graphite

Carbon.

Volatile.

Per

Cent.

99-79

.16

66.40

.70

79.40

5.10

Ash.

Per

Cent.

.05
32.90
15.50

Bauxite was at one time much experimented with for a
neutral lininpf, but it has practically no use at this time in
steel making. The excessive shrinkage and difficulty in

§ 35 MANUFACTURE OF STEEL 45

making linings caused it to be given up. It is a hydrated
oxide of alumina, but seldom occurs without the admixture
of more or less oxide of iron. Bricks for other uses are
made of it to a limited extent.

GrapMte, in the strictest sense, is the only neutral refrac-
tory, as the other two can be made unite with other elements
under the right chemical and heat conditions. It is used in
making crucibles and rarely mixed with ball stuff. Bricks
have been made of it for the hearth of blast furnaces, but
they are not suitable for open-hearth furnaces owing to the
oxidizing conditions destroying them and the metal readily
absorbing the carbon.

Table XI shows analyses of neutral refractories.

■

62. Refractory Mixtures. — Almost every steel works
has its own particular mixtures for different purposes, but
in general the following are standard practice for the uses
shown. The proportions of all will be varied according to
results obtained and quality of materials used.

1. Bottom stuffs for making bottoms of acid Bessemer '
converters : Ganister (or quartz), 15 per cent. ; fine sand
(high SiO^)^ 25 per cent. ; clay, 25 per cent. ; coke dust
(aids drying), 15 per cent. ; ground clay bricks, 20 per cent,
(mostly bats).

2. Cupola stuffs for patching cupolas, iron troughs, etc. :
Ganister (or quartz), 50 per cent. ; sand, 25 per cent. ; clay,
25 per cent

3. Vessel patchings for putting on bottom and patching
nose of acid converter: Ganister (or quartz), 68 per cent.;
sand, 16 per cent. ; clay, 16 per cent.

4. Ladle lining and patching : Either loam or vessel
patching.

63, Other works use one mixture for most of the above
purposes — one of about half ganister and half clay is com-
monly used for everything except ladles, loam being almost
always used for patching these. Large ladles are always

46 MANUFACTURE OF STEEL § 35

bricked up, while smaller ones (for Bessemer and small
open-hearth furnaces) are lined with ball stuff of clay and
ganister. It must be remembered that, just as in the fur-
nace or vessel, the lining must correspond to the slag,
although greater variations between them are allowable in
the ladle, etc., as the slag is exposed to the air and partially
chills, so that the reaction is much less vigorous. It would
not do to use ganister to line a ladle for the basic, or a clay
or loam with much free lime for the acid process; most loam
and clay are, however, sufficiently neutral to be used for
either.

INDEX

. Note.— All items in this index refer first to the section (see the Preface) and then
on the page of the section. Thus, '* Alloy steels 84 56" means that alloy steels wiU
be found on page 56 of section 34.

A Sec. Page

Absorption chambers for

bleaching powder 30 30

Acid and basic open-hearth

systems £8 23

Bessemer converter, Lin-
ingof 84 14

Bessemer process 84 13

Bessemer process. Chem-
ical changes in the con-
verter of .' 84 16

Bessemer process. Gen-
eral arrangement of
plantof 34 19

Bessemer process. Tem-
perature in converter
of 34 17

free, Determination of,
in salt cake 81 19

free, Determination of,
instill liquor 81 43

Mixed 28 68

Nordhausen or fuming
sulphuric 97 4

open-hearth process. ... 33 44

open-hearth process, Ad-
dition of ore in 83 55

open-hearth process,
Charge of 88 45

open-hearth process.
Finishing the heat in.. 88 68

open-hearth process, Re-
moval of metalloids in 83 54

open-hearth process,
Slagin S3 53

pumps 28 25

refractories 85 88

sulphuric, Definition of. 27 8

»t

Sec. Page
Acker process for the electrol-
ysis of salt 30 71

Afterblow of basic Bessemer

process ,34 25

Air-quenched steel 84 56

** valves for open-hearth fur-
nace 33 19

Alkali, total. Determination of,

in black ash 81 22

total, Determination of,

in caustic bottoms.... 31 35
total. Determination of,

in caustic liquor 81 84

total. Determination of ,

in caustic mud 31 85

total, Determination of,

in caustic soda 31 86

total. Determination of,

in fished salts 81 34

total. Determination of,
in lye from extraction

of black ash 31 23

total. Determination of,

in soda ash 31 t!8

Alkaline sodium compounds.
Determination of, in tank

waste 81 27

Alloysteels 84 56

Alumina, Determination of, in
lye from extraction

of black ash 31 24

Determination of, in

salt cake 31 19

Determination of, in

soda ash 31 16

in brine. Determina-
tion of . 31 %

44'

INt)BX

14
6

Sec. Pag§

Aluminum bleach 80 80

" Effect of, in 8teel... 85 14

Use of, in basic 8teel 84 48

Ammeters 80 K5

Ammonia, Determination of, in

ammonia liquor.. 81 7
" Determination of, in

ammoniacal brine 81 7
** Determination of, in
bicarbonate from

filters 81

'* Determination of, in

mother liquor 81

** liquor, Analysis of.. 81
*' liquor. Determina-
tion of ammonia in 81 7
** liquor. Determina-
tion of specific

gravity of ... 81 6

** lost in Solvay proc-
ess 89 81

** recovery in Solvay

process 80 28

** soda. Analysis of ... 81 1
*' soda, Properties of.. 80 81
'* used in Solvay proc-
ess 20 14

Ammoniacal brine 89 80

" brine, Analysis of 81 7

'* brine, Carbona-

ting.... SO 21

" brine, Determina-
tion of ammo-
nia in 81 7

** brine. Determina-

tion of salt in... 31 7

Anions 80 60

Anode 80 60

Anthracite for blast furnace... 82 16
Appparatus employed in the
chamber process
forsulphuricacid 28 5
*' used in Bessemer

process 84 8

Area reduction test of steel.... 86 21
Arsenic, Determination of, in

hydrochloric acid... 81 39
'* Freiberg process for
removing, from

chamber acid 28 _48

" in iron 32 60

" Precipitation of, in
the Freiberg proc-
ess 28 46

** Purification of cham-
ber acid from 88 48

Sec, Page
Arsenic, Stahl method for re-
moving, from cham-
ber acid 88 47

Available sulphur 87 15

** sulphur in burner

gas 87 14

B . Sec. Page

BarilU 80 10

Barium chlorate 80 48

Basic Bessemer process 84 80

" Bessemer process, Action

of basic fluxes in 84 80

** Bessemer process. Blow-
ing the charge 84 85

•* Bessemer process, Con-
trol of blowing in 84 86

** Bessemer process. Oxi-
dation of the elements in 84 87
" Bessemer process, Pig

iron used in 84 23

" Bessemer slags as fertil-
izers 84 81

** converter 84 28

** fluxes. Action of, in basic

Bessemer process 34 '80

** hearth materials 33 68

" open-hearth charge 88 64

*' open-hearth charge. Cal-
culation of 33 66

*' open-hearth furnace,

Thermal conditions in.. 88 79

** open -hearth process 83 60

** open-hearth process, Ad-
vantages of 88 61

** open-hearth process, De-

phdsphorisation in 88 79

** open-hearth process, £>e-

sulphurization in 88 81

*' open-hearth process,

Meltingin 83 73

•* open-hearth process, Use

of ore in 88 71

** open-hearth slag 88 74

*^ refractories 35 43

** steel, Use of aluminum in 84 48
Baumd specific gravity scale,

European 87 6

** specific gravity scale.

United States 87 6

Bauxite 85 44

** as hearth material 88 83

Bell of blast furnace 82 88

Bertrand-Thiel process for
making steel 85 80

Bi58§«^|^r(;opv^rt^^^..... ..•••• ^H W

INDEX

Sec.

Bessemer process 84

** process. Acid 84

** process, Apparatus

used in 84

** process. Blast for — 84
^ process. Blowing en-

g^ine for 84

*' process. Casting in.. 84
** process, Cupolas for. 84
Bicarbonate from filters. Anal-
ysis of 81

** from filters, De-

termination of

ammonia in 81

•• from filters. De-

termination of

moisture in 81

*• from filters. De-
termination of
sodium bicarbo-
nate in 31

•• from filters. De-

termination of
toUl alkali in... 81

Billet 85

Bittern »

Black ash 29

** ash, Analysis of 81

** ash. Analysis of lye from

extraction of 81

** ash, Composition of 29

** ash. Cyanides in 29

** ash. Determination of

caustic soda in 81

^ ash. Determination of

free lime in t.... 31

'* ash. Determination of

salt in 81

** ash. Determination of so-
dium carbonate in 81

** ash. Determination of so-
dium sulphate in 81

** ash, Determination of

total alkali in • 81

" ash, Lixiviation of 29

** ash. Properties of 29

" ash. Sampling of 81

Blanquette 29

Blast for Bessemer process 84

•* for blast furnace. Practi-
cal suggestions for... . 82

•* furnace 82

** furnace. Bell of 32

** furnace. Calculation of

burdens for 82

** furnace, Charging of.... 42

Pasre

1
16

48
21

28
M

22
57
66

21
10
16

61
26
82

48
85

Sec, Page
Blast furnace. Hanging and

slipping in 82 65

** furnace, Hopper of 83 32

** furnace, Miscellaneous
matters concerning

operation of 83 66

" furnace. Reactions in ... . 83 40

*^ furnace. Reactions in ... . 32 44

** furnace. Scaffolds in 82 64

** furnace slag, Composi-
tion of 82 45

** furnace slags, Fusibility

of 82 46

** furnace tuyeres 82 29

Bleach, Electrolytic 80 83

** liquors, Analysis of.... 81 60
** liquors. Determination
of available chlorine

in 81 48

** liquors. Determination

of carbonates in 81 49

^ liquors. Determination

of caustic alkali in... 81 49
** liquors. Determination

of chlorates in 81 48

** liquors, Determinatipn

of chlorides in 81 48

** Valuation of 80 88

Bleaching powder 80 29

** powder, Absorption

chambers for 80 80

** powder, Analysis of 81 60
•* powder. Chlorine for

making 80 81

** powder, Composi-
tion of 30 33

** powder. Lime for

making SO 80

** powder, Properties of 30 32

" powder. Uses of 30 35

Blind roaster for salt cake 29 41

Bloom 3-) 34

Blooming mill.... &5 34

Blowholes in steel castings 85 9

** in steel. Prevention

of 85 11

Blowing basic Bessemer charge 34 25
•* engine for Bessemer

process 34 16

" engines 32 18

" in the blast furnace... 32 35
** out the blast furnace. 32 88
Blumenberg electrolytic proc-
ess for potassium chlorate. . . 80 86
Bombonnes for condensing hy-
drochloric acid 80 4

INDEX

Sec.
Bottom of acid Bessemer con-
verter .;.. 84

** of basic converter 84

'* stuff 85

Breaking throngh of flame 84

Bridging in lime kiln 29

Brimstone S7

" burner, Harrison-
Blair 27

'* burners 27

Brine, Ammoniacal 29

*' ammoniacal. Analysis of 81
** ammoniacal. Analysis of 81
" ammoniacal. Determi-
nation of ammonia in.. 81
** ammoniacal. Determi-
nation of salt in 81

** Determination of cal-
cium oxide in 81

*' Determination of ferric

oxide and alumina in. 81
** Determination of inor-
ganic sediment in 81

*' Determination of mag-
nesia in 81

** Determination of so-
dium chloride in 31

** Determination of spe-
cific gravity of 81

** Determination of sul-
phur trioxide in 81

** Evaporation of, by

grainers 29

" forSolvay process, Puri-
fication of 29

" Kettle evaporation of . . 29
" Pan process for evapo-
ration of 29

" Salt from 29

" Solar evaporation of 29

^* used in Solvay process. 29
" Vacuum pan process for

the evaporation of ... . 29

Brown hematite 82

** hematite. Distribution

of, in United States. . . 82

Buckstays 83

Bunte, Burette 81

'* Burette, Reagents for.. 81
Burdens for blast furnace, Cal-
culation of 8S

Burette, Bunte 81

Burned ball 29

Burner gas 27

" gas. Available sulphur

in 27

Pajre

14
iZ
45
17
18
8

26
25
20

20
5

6
4

4
14

6
4

Sec.
Burner gas. Calculation of vol-
ume of 27

** gas. Collecting sample

of 27

" gas, Furnaces and
burners for the pro-
duction of 27

** gas, Production of 27

** gas. Purification •of. ... 27

** gas. Reheating of 27

** gas, Reich^s test for sul-
phur dioxide in 27

" gas. Testing 27

** Harrison-Blair brim-
stone 27

Burners and furnaces for the
production of burn-
er gas 27

*• Brimstone.. 27

Pyrites 27

C Sec.

Calcination of iron ores 82

*' of sodium bicar-
bonate 29

Calcining furnace for sodium

bicarbonate 29

" kiln, Gjers 82

*' soda crystals 29

Calcium carbonate. Determina-
tion of, in quicklime. 81
** carbonate. Determina-
tion of, in soda ash... 81
*' carbonate for Le Blanc

soda process 29

** chloride, Determina-
tion of, in still liquor 81
** fluoride from cryolite

soda process 29

** oxide. Determination

of, in brine 81

" oxide, free. Determi-
nation of, in quick-.

lime 81

Calorific value of producer

gas 38

Carbide of silicon 84

Carbon as hearth material ... 83
" Control of, in recarbon-

izing 84

** dioxide. Determination

of, in manganese ore 81
** dioxide. Determination

of, ia slaked lime 81

*' dioxide for Solvay

process . . 29

Pag^e
40
88

23
21
50
58

86
86

25
27

Pa^e
9

27
10
68

6
16
48
44
84

85
02

INDEX

xiii

Sec.
Carbon dioxide for S o 1 v a y

process. Washing of. 89

*' Effect of, in steel 85-

* * for LeBlanc soda process 89

" iniron 88

'* Oxidation of, in basic

Bessemer process. ... 84
Carbonate iron ore. Distribu-
tion of, in United States. 88

Carbonates, Determination of,

in bleach liquors 81

Carbonated lye. Analysis of... 81
** lye, Determination
of, sodium bicar-
bonate in 81

Car bonatinganimoniacal brine 89
** tower for Solvay

process 89

Carborundum 84

Casting from blast furnace.... 88
** in Bessemer process. . . 84
** steel. Composition of.. 85
** steel, Tropenas process

for 84

Castings, Steel 8S

CaHtner-Kellner electrolytic

process for salt 80

Catalytic or contact process for
the manufacture of sulphuric

acid 87

Cathode 80

Cations 80

Caustic alkali, Determination

of, in bleach liquors. SI

" bottom 89

** bottonis, Analysis of.. 81
**• bottoms. Determina-
tion of insoluble mat-
ter in 81

*^ bottoms, Determina-
tion of salt in 81

'* bottoms, Determina-
ation of sodium car-
bonate in 81

** bottoms. Determina-
tion of total alkali in 31
*■* lime. Determination

of, in caustic mud... 81
*'*' liquor. Analysis of.... 31
*' liquor. Determination

ofsaltin 31

'* liquor. Determination

of specific gravity of 81
^* liquor. Determination
of total alkali and so-
di um carbonate in. . . 31

Pagre

18
18
48
55

49
84

85
81

88
85
89
19
8

81
1

48
60
60

49
81
84

85
85

85
38

Sec,
Caustic liquor. Filtration of... 89

** mud 29

** mud. Analysis of 81

** mud. Determination of

calcium carbonate in 31
** mud. Determination of

total alkali in 81

" pots 89

** soda. Analysis of 81

**" soda, crude materials

for. Analysis of 81

** soda. Determination

of, in black ash 81

** soda, Determination
of, inlye'f rom extrac-
tion of black ash 81

** soda. Determination

of, in soda ash 81

** soda. Determination

of total alkali in 81

** soda. Uses of 89

Causticizing sodium carbonate 89
Centrifugal ventilator for Sol-
vay process 89

Chamber acid, Freiberg proc-
ess for removing

arsenic from 88

acid. Purification of. 88
** acid, Purification of,

from arsenic 88

*' acid, Stahl method
for removing arsen-
ic from 28

crysUls 88

" process. Diagram of. 28
^* process for sulphuric

acid 88

** process for sulphuric

acid. Control of . . . . 28
*^ process for sulphuric

acid. Operation of. 88
** process for sulphuric

acid. Reactions of. 88
*' process for sulphuric

acid, Starting of. . 88

Chambers, Lead 38

Chance-Claus process for re-
covery of sulphur from tank

waste 89

Charcoal for blast furnace 38

Charge, crucible, Melting of. . . 84
** for basic open-hearth

furnace 33

** for basic open-hearth
furnace. Calculation
of 88

Page
77
88
85

85
80
86

86
83
76

48
43

1
38
28

85
11

69
14
50

xiv

INDEX

Sec,

Chaige for crucible 84

*' o£ acid open-hearth
furnacct Calculation

of 88

** of acid open-hearth

furnace, Melting the 83
Charging acid oi>en-hearth fur-
nace, Method of ... . 83
*' machine, Wellman... 88

" of blast furnace 32

Chemical changes in the con-
verter of acid Bes-
semer process 84

** composition of steel.
Relation of, to

strength 85

*^ examination of steel 85
Chlorate, potassium. Analysis

of 81

Chlorates, Analysis of. 81

** Determination of, in

bleach liquors 81

Chloride of lime 80

Chlorides, Determination of, in

bleach liquors 81

Chlorine 80

^^ available. Determina-
tion of, in bleach

liquors 81

** by nitric-acid process 80
" by the Weldon proc-
ess 80

** Deacon's process for.. 80

" direct from salt. 80

** for making bleaching

powder 80

** from hydrochloric

acid 80

Liquid 80

♦* Sourceof 30

Chrome steel 84

Chromite 85

'^ as hearth material... 88

Cinder notch 88

Claus kiln for sul phur recovery 29

Clay for making crucibles 84

" ironstone 82

Clays 85

Clear liquor. Determination of

calcium chloride in 31

Coal, Analysis of 81

'* used in Solvay process... 29

Coke, Analysis of 81

'* for blast furnace 82

** tower for condensing

hydrochloric acid 80

4«
25
85

22
16

50
60

48
29

48
8

48
28

10
29

9
60
44
62
81
71
47

5
88

7
15

7
14

Sec.
Coke used in Solvay process.. 29

Cold shortness 85

Combustion of sulphur 27

Commercial methods for deter-
mining the strength of sul-
phuric-acid solutions weaker

than the monohydrate 27

Concentrating pots for the
manufacture of potassium

chlorate 80

Concentration and distillation
of sulphuric
acid starting
with the Glov-

er*8 tower 28

^ of dilute sul-
phuric acid so-
lutions 28

^ of sulphuric
acid by the
Kessler proc-
ess 28

** of sulphuric

acid in glass
retorts or stills 28
** of sulphuric

acid in iron.... 28
** of sulphuric

acid in lead

pans 28

^ of sul phuric
acid in plati-
num «... 28

** ofsulphuric

acid in porce-
lain or glass
beakers or

dishes 28

Condensation of hydrochloric

acid 80

Condenser, Paid ing 28

" Gilchrist 28

Lunge 28

Condensers, Surface 26

Conditions in the chambers in
the manufacture
of sulphuric acid. 28
** in the Glover's

tower 28

Conductivity of solutions 80

** of solutions, De-
termination of. 80
*♦ of solutions. The
effect of tem-
perature on.... 80
vessel 80

Pagre
16
18
12

52
51

19
18
16
16

29
49

49
50

INDBX

Sec,

Conductors, Electric 80

Constant-temperature bath. ... 80
Contact mass or material used
in the manufacture
of sulphuric acid in

the contact process . . S7

** ovens 27

** process for sulphuric

acid, Diag^ram of . . . . in
** process for the manu-
facture of sulphuric

aQid 87

Control of blowing in basic

Bessemer process 84

Converter, Basic 84

** Bessemer 84

^ for acid Bessemer

process. Bottom

and tuyeres of . . . . 84

" Prasch 27

•» Tropenas 84

Copper, Effect o^ ^n steel 85

'* in iron 82

'* nickel pyrrhotites 27

Copperas slate 27

Cor bin electrolytic process for

potassium chlorate 80

Cowper stove 82

Cranes 83

Crook'sstove 82

Crucible charge 84

" charge, Melting of.... 84

** furnace 84

** process 84

** steel. Composition of. 84
** steel. Superiority of.. 84
Crucibles for steel manufac-
ture 84

** Manufacture of 84

** Materials for manu-
facture of 84

" Teeming or pouring

of.. 84

Crushing iron ores 82

Cryolite soda process 29

" soda process. Calcium

fluoride from 29

** soda process. Sodium

aluminate from 29

Crystal soda. Analysis of 81

Crystallizing pans for the man-
ufacture of potassium chlo-
rate 80

Cupola, Flux for 34

" Fuelfor 84

•* lining 84

59
50

45
58

20
28

14
40
81
14
00
8
8

86
21
26
25
49
60
44
44
54
58

84
29

41
7
6
6

Sec.

Cupola mix 84

" stuff 86

Cupolas for Bessemer process. 84

*' Tuyeres for 84

Curtain drip 28

Cut-off gas valves 83

Cyanides in black ash 29

D Sec.
Dampers for open-hearth fur-
nace 83

Darby method of recarbonizing 84

Davis-Colby ore roaster 82

Deacon-Hasenclever process
for the purification of hydro*

chloricacid 80

Deacon's plus-pressure.f umace

for salt cake 29

** process for chlorine.. 80

Dead melting steel 84

Denaturated salt 29

Density of sulphuric acid. De-
termination of 27

Dephosphorization in basic

open-hearth process 83

Desulphurization in basic open-
hearth proc-
ess 33

** in basic open-

hearth proc-
ess, Saniter

Process for. 33

Desulphurizing ores. Rules for 32

Diagram of chamber process. . 28
" of contact process for

sulphuric acid 27

** of manufacture of

sulphuric acid 28

Dissociation, Electroljrtic 80

Distiller liquor in Solvay proc-
ess 29

Dolamite 85

** as hearth material ... 33

Downcomer 32

Dropofflame 84

Duplex process for making steel 35

Dust catcher 32

B Sec.

Eaudejavel 80

** dejavel 80

** de LabaVraque 80

Elastic-limit test of steel 85

Electric cond uctors 80

*' current. Sources of.... 80

** polarization 80

7
45
4
5
40
22
64

28
41
18

83
14
33

60
62

80
43
63
32
17
3;!
32

8
35
35
19
59
44
68

ZVl

INDBX

Sec,

Electricity, Quantity ot 80

Electrodes 80

** for the electrolysis

of salt 80

Platinizing 80

Electrolysis 80

** ofsalt 80

" of salt by Castner-

Kellner process.. 80
" of salt by Green-

wood process. ... 80
" of salt by Har-

grrea ves^nd>Bi rd

process 80

" of salt by Hulin*s

process 80

** of salt by processes

nsing: a mercury

cathode 80

** of salt by proc-
esses using dia-
phragms 80

** of salt by the

Acker process . 80
** of salt by the Le

Sueur process. . . 80
^* of salt. Conditions

favoring 80

"* of salt with dis-
solved electro-
lyte... 80

** of salt with fused

electrolyte 80

Electrolytes 80

Electrolytic bleach 80

'* dissociation 80

** methods for the

production of al-
kali and chlorine 80
** potassiumchlorate 80

Electromotive force 80

** force. Measure-
ment of 80

** force of polari-
zation, Calcu-
lation of, from
heat of reac-
tion SO

Elongation test of steel 85

Engines, Blowing 8*2

Evaporating pans with heat be-
low 59

Evaporation of brine by grain-

ers 29

'* of brine by vacu-

um pan process. S9

Pa/re
M
60

68
51
60
66

T7
€9

74
71
75
67

60
60
8^
62

84
56

80
18

Sec,

Evaporation of brine, Kettle. . . 89
** of brine, Pan proc-
ess for 29

** of brine, Solar. ... 89

** of sodium hydrate 89

*' of tank liquor.... 29

Evaporator, Yaryan 89

Exit gas. Determination of per-
centage of sulphur dioxide in 87
Explosion doors of blast fur-
nace 88

P Sec,

Raiding condenser 88

" lump burner 87

Faraday's law 80

Ferric oxide. Determination of,
in lye from extraction

of black ash 81

** oxide. Determination of,

in salt cake 81

** oxide. Determination of,

in soda ash 81

** oxide in brine. Deter-
mination of 81

Ferromanganese 88

Ferrosilicon 84

Ferrous carbonate 82

Fertilizers, Basic Bessemer

slags as 34

Filter for filtering caustic

liquor 89

Filters for Sol vay process. .... 29

Filtration of caustic liquor. ... 89

Fished salts. Analysis of 81

** salts. Determination of
oxidizable compounds

in 81

** salts. Determination of

sodium sulphate in . . . 81
'* salts. Determination of

toUl alkali in 81

Flux for cupola 84

Fluxes, basic. Action of ,in basic

Bessemer process 84

" for iron ores 88

Foaming in basic open-hearth

furnace 83

Foreblow of basic Bessemer

process 84

Forehearth of Wellman rolling

f urnace 88

Forter valve 88

** water-seal producer.... 88

Franklinile 38

Page
5

6
4

78
60
78

Pasre
19
27
60

19
16

8
59
84
84

77
84
77
34

34
7

80
16

INDEX

xvii

Sec.

Prasch converter 27

Praser-Talbot mechanical pro-
ducer 83

Free acid. Determination of, in

salt cake 81

** lime, Determination of, in

black ash 31

Freezing^ process for the pro-
duction of sulphuric mono-
hydrate, Lungre 88

Freiberg process for removing
arsenic from cham-
ber acid 28

** sulphureted hydro-
gen generator 28

Fuel for blast furnace 82

** for cupola 84

** for making producer

gas 88

" Gaseous, used in open-
hearth furnaces 38

Fuming sulphuric acid 27

Furnace, Blast 82

Crucible 84

** for calcining sodium

bicarbonate 29

*^ for salt cake, Dea-
con's plus-pressure. 29
*' Herreshoff, of the

MacDougall type... 27

** Maletra-Falding 27

** muffled, MacDougall

type of 27

** Open-hearth 83

** open-hearth. Con-
struction of PS

** pit 85

** Rhenania muffled

type of. 27

*' rolling, Advantages

of 88

" Spence, Reciproca-
ting type of 27

" Wellman rolling 88

Furnaces and burners for the
production of burn-
er gas 27

•* for Le Blanc process.

Mechanical 29

** for Le Blanc soda

process 29

** for salt cake. Me-
chanical 29

" open-hearth. Capac-
ity of 88

*♦ Reheating 86

Paj^e Q Sec. Pagre

49 Gall-and-Montlaur process for

potassium chlorate 30 85

86 Gas, Artificial, as fuel for crpen-

heatth tumace 88 82

19 *» Burner 27 18

'* from bleaching -powder
21 chambers, Testing of . . . . 81 47
** from decomposer, Anal-
ysis of 81 40

59 ** from gasometer. Analysis

of 81 80

** from sulphate pan, Anal-

48 ysisof 81 45

'* Natural, as fuel for open-

44 hearth furnace 88 80

14 *' Natural, Composition of. . 88 81
6 " Producer 83 88

** valves for open-hearth

88 furnace 83 19

Gaseous fuel used in open-

80 hearth furnaces \ 88 80

4 Gases from ammoniasaturator,

26 Washing 29 24

44 *' lime-kiln. Analysis of.... 81 80

Gay-Lussac tower 28 21

27 ** Lussac towers. Number of 28 87
Gerster's formula 27 21

41 Gibb's electrolytic process for

potassium chlorate 30 87

80 Gilchrist condenser 28 18

80 Gjers calcining kiln 82 lO

Glauber's salt 29 86

84 Glover's tower 28 8

6 " tower. Conditions in.. 28 29
Grading of crude sulphur 27 9

7 Gt-ainers for evaporating brine 29 6
33 Graphite 86 46

** for making crucibles. 84 46
84 Greenwood process for the elec-
trolysis of salt 80 74

17 Grinding crystals of potassium

chlorate 80 42

81 " sodaash 29 63

15 Grittinger ore roaster 82 12

H Sec. Page

28 Hadfield's manganese steel .... 84 69
Hanging in blast furnace 82 65

61 Hargreaves-and-Bird process

for the electrolysis of salt. ... 80 77

49 Harrison-Blair brimstone

burner 27 26

44 Hart system for absorption of

hydrochloric acid 80 5

18 Hasenclevor method for puri-

83 fication of hydrochloric acid. 80 7

3cviii

INDEX

Sec.

Hearth materials, Basic 88

" materials. Neutral 88

*' of open-hearth furnace 88

Hematite, Brown 8S

" brown, Distribution

of, in United States 88

** Red 8S

** red, Distribution of,

in United SUtes.. 82
Herreshoff furnace of the Mac-

Dougall type 27

Honeycomb in ing^ots 85

Hopper of blast furnace 82

HuUn*8 process for the electrol-

ysisofsalt 80

Hydrates and solutions of sul-
phur trioxide 27

** of sulphur trioxide.

Nomenclature of.. 27
Hydrochloric acid. Analysis of 81
** acid. Analysis of

finished product 81
** acid, Analysis of

waste gases
from absorption

of 81

** acid. Apparatus
used for con-
densing 80

** acid, Commercial 80

** acid, Condensa-

tionof 80

" acid, Determina-

tion of arsenic in 81
** acid. Determina-

tion of hydro-
chloric acid in.. 81
'* acid, Determina-

tion of selenium

in 81

•* acid. Determina-
tion of sulphur-
ic acid in 81

** acid, Determina-
tion of sulphur-
ous acid in 81

** acid, Oxidation

of, by oxides of

manganese 80

•* acid, Process of

manufacture of 80
*• acid. Purification

of 80

*• acid. Purification
of, by the Dea-
con-Hasenclev-
er process 80

Page

8
4

80
10
82

3
36

8
6

10
1
7

See. Page
Hydrochloric acid. Qualitative
tests for arsenic

in 81 40

•* acid. Uses of 80 8

Hydrogen sulphide 27 8

** sulphide generator,

Freiberg 28 44

*• Sulphureted 27 10

I Sec. Page

Ingot, Treatment of 85 82

Inorganic sediment in brine.

Determination of 31 2

Insoluble matter. Determina-
tion of, in caustic

bottoms 81 85

** matter in limestone.

Determination of. . 81 4
'^ matter in quicklime.

Determination of . . 81 6

Ions 80 60

** Migration velocity of 80 02

Iron, Arsenic in 82 60

" Carbon in 82 66

•• Copper in 82 60

** Elements contained in.... 33 66

'* Grading of, by analysis. . 82 54

" Grading of, by fracture.. 32 64

'* Manganese in 32 66

" notch 82 81

" notch. Trouble with 82 65

" ore. Valuation of 32 7

** ores. Classification of 82 8

** ores. Classification of.... S2 68

'* ores, Crushing 82 8

** ores. Distribution of, in

the United States 82 5

** ores. Preparation of 82 7

** ores. Roasting 82 9

** ores, Washing 82 7

** Phosphorus in 82 67

" pyrites 27 8

♦* Silicon in 82 57

" Sulphurin 82 69

'* Titanium in 82 60

K Sec. Page

Kennedy*s stove 82 25

Kessler process. Concentration

of sulphuric acid by. « S8 55

Kestner automatic pump 28 25

Kettle evaporation of brine.... 29 5

Killing steel 84 61

L Sec. Page

Ladle lining and patching 85 45

^* steel 88 27

INDEX

XIX

Sec.

Lead chambers 88

^* chambers, Admission of

steam to 28

** chambers. Conditions in. 88
L« Blanc process. Advantages
and disadvantages
of mechanical fur-
nace in •••••••••••••• tb§

^* process, Management
of mechanical fur-
nace in 89

** process, Mechanical

furnaces for 29

•* process. Waste in ... . 89

** soda process 89

** soda process, Calci u m

carbonate for 89

** soda process. Carbon

for 89

^ soda process, Details

of 89

** soda process. Fur-
naces for 89

** soda process. Hand

furnaces for 89

** soda process. Man-
agement of furnace

for 89

** soda process, Raw

materials for 89

Le Sueur process for the elec-
trolysis of salt 80

Lime, Addition of, in basic

open-hearth process.. 88

'* as hearth material 88

** caustic. Determination

of, in caustic mud.... 81
** Determination of, in

limestone 8t

" Determination of, in salt

cake 81

" free, Determination of,

in black ash 81

** kiln for making carbon
dioxide for Solvay
process .........i •.... <vV

*' kiln gases. Analysis of. 81
** kiln gases. Analysis of. 81

" milk of. Analysis of 81

" milk of. Determination

of specific gravity of. 81

** slaked. Analysis of 81

** total. Determination of,

in black ash 81

** used in making bleach-
ing powder 30

88
80

58
51

4«

48
48
48
49
49

69
83

80
21

14
48

Sec, Page
Lime used in the manufacture

of sodium hydrate 89 76

Limestone 85 48

** Analysis of 81 4

** Determination of in-
soluble matter in. 81 4
" Determination o f

lime in 81 5

** Determination o f

magnesia in 81 5

*' EflBcIency of, as

flux 88 17

** used in Solvay proc-
ess 89 18

Limonite 88 4

*' Distribution of, in

United States 88 8

Lining for cupola 84 6

^* of acid Bessemer con-
verter 84 14

^* of basic converter 84 83

Linings, Acid and basic, for

open-hearth furnace 88 85
** Protection of blast-
furnace 88 88

Liquid chlorine 80 89

Liquor, caustic. Analysis of.... 81 83
'* from cart>onators, An-
alysis of 81 11

Lixi viation of black ash 89 57

Loewig*s process for manufac-
ture of sodium hydrate 89 83

Lump burner, Falding 87 27

Lunge condenser . . 38 18

** freezing process for the
production of sulphu-
ric monohyd rate 88 59

** plate column 28 18

'* plate tower for condens-
ing hydrochloric acid.. 30 5
Lye, carbonated, Analysis of . . 31 84
*^ from black ash. Analysis

of 81 28

** from black ash. Determi-
nation of caustic soda in 81 88
** from black ash. Determi-
nation of salt in 81 88

** from black ash. Determi-
nation of silica, ferric
oxide, and alumina in . . . 81 84
** from black ash. Determi-
nation of sodium carbo-
nate in 81 98

*' from black ash. Determi-
nation of sodium ferro-
cyanide in 81 84

INDEX

Sec, Page
Lye from black ash. Determi-
nation o£ sodium sul-
phate in 81 14

** from black ash, Determi-
nation of sodium sul-
phidein 81 28

*' from black ash. Determi-
nation of specific gravity
of 81 28

*' from black ash. Determi-
nation of total alkali in.. 31 28

'^ from black ash. Determi-
nation of total sulphur in 81 24

" Purification of 29 S»

M Sec. Page

MacDougall type of muffled

furnace 27 84

Magnesia 85 42

" Determination of, in

brine 81 8

'' Determination of, in

limestone 81 5

'* Determination of, in

quicklime.... 81 6

'* Determination of, in

salt cake 81 20

Magnesium carbonate, Deter-
mination of, in soda ash 81 17

Magnesite as hearth material.. 88 08

Magnetite 82 4

'* Distribution of, in

United States 82 6

Maletra-Palding furnace 27 80

Manganese, Effect of, in steel . 85 18

iniron 82 58

in Weldon mud.
Determination of 81 44
** Loss of, in recar-

bonizing 84 80

^* ore, Analysis of... 81 41
*' ore, Determina-
tion of acid nec-
essary to decom-
pose 81 48

*' • ore. Determina-
tion of available

oxygen in 81 42

/' ore, Determina-
tion of carbon di-
oxide in 81 48

" ore. Determina-
tion of moisture in 81 41
*' Oxidation of, in
basic Bessemer
process 84 27

Sec. Page

Manganese steel 84 58

" Weldon *s process

for the recovery
of, from still

liquors 80

Massick stove 82

Mechanical furnaces for salt

cake

" pans for the evapo-

ration of tank

liquor 20

Metalloids, Removal of, in

open-hearth acid process 88

Microscopical examination of

steel 85

Migration velocity of ions 80

Milk of lime. Analysis of 81

*' of lime, Specific gravity of 81

Mill, Blooming 85

•* Plate 85

•» Slabbing 85

" Universal 85

** Universal plate 85

Mills, Rolling 85

" Steel 85

Mirabilite 29

Mixedacid 28

Mixer for Bessemer process.. .. 84
Moisture, Determination of, in
bicarbonate from

filters 81

** Determination of, in

manganese ore 31

Monell steel process 85

Monohydrate of sulphur triox-

ide 27

Monte jus's pump with acid egg 28

Mother liquor, Analysis of 81

" liquor, Determination

of ammonia in 81

*' liquor. Determination

of salt in i... 81

Mud, Caustic 29

^* caustic. Analysis of 81

♦' Weldon 80

** Weldon, Analysis of 81

Muffle^ roaster for salt cake .... 20
Muffled furnace, MacDougall

type of 27

** type of furnace, Rhe-

nania 27

N Sec. Page
Natural gas as fuel for open-
hearth furnace 83 80

** gas, Composition of.. 83 81

16
25

29 44

15
62
14
14
84
86
87
87
87
84
88
85
68
9

41
29

8
26
18

14
82
85
19
44
41

INDEX

XXI

Sec. Page
Natural sras. Introduction of,

into furnace 83 88

Neutral hearth materials 38 08

** refractories 85 48

Nickel, Effect of, m steel 85 14

♦• steel 84 50

Niter oven. Definition of. 87 i5

Nitrating: by potting: *8 5

by use of nitric acid. 28 7

oven 88 5

Nitric-acid chlorine process... 80 88

Nitrous vitriol. .. .. SB 5

Nitrosulphuric acid ^ 8

Nomenclature of solutions av.d

hydrates of sulphur trioxide 87 8

Nordhausen sulphuric acid. ... 27 4

O Sec. Page

Ohm'slaw 80 50

Open-hearth furnace 83 6

hearth furnace, Artificial

gas as fuel for 88 88

hearth furnace, acid, Cal-
culation of charge of... 33 47

hearth furnace. Capacity

of 83 18

hearth furnace, Charge

ofacid 83 52

hearth furnace. Construc-
tion of 83 7

hearth furnace, Gas and

air valves for 33 10

hearth furnace. Hearth

ofacid 83 44

hearth furnace, Method

ofcharging acid 83 46

hearth furnace, Method

of heating acid 83 58

hearth furnace, Ports to. 33 13

hearth furnace. Roof of.. 33 7

hearth furnaces. Gaseous

fuel used in 83 30

hearth process, Acid 33 44

hearth process. Charge

ofacid 33 45

hearth process for steel.. 83 5

hearth process, Talbot

continuous 35 85

roasters for salt cake 20 86

Ore, Addition of, in acid open-
hearth process 33 55

Definition of 82 8

Reduction of, in blast fur-
nace S2 48

Use of, in b a s i c o p e n -

hearth process 38 71

Sec. Page

Ore, Valuation of iron 88 7

Ores, Classification of iron 88 8

Oven, Contact 87 58

Nitrating 88 5

Oxides, Effect of, in steel 85 14

Oxygen, available. Determina-
tion of, in manga<

nese ore 81 48

**' Effect of, m steel 85 14

P Sec, Page
Pan process for the evapora-
tion of brine 80 5

Pauli's method for purification

oflye 80 60

Pechiney-Weldon method for
removing cyanides from

black ash 80 54

Phosphorus, Effect of, in steel. 85 18

in iron 88 87

*' Oxidation of, in

basic Bessemer

process.....* 84 88

Physical testing of steel •. . 85 16

** testing of steel. Re-
sults in 85 85

Pig iron used in basic Bes&emer

process 84 82

Pigs 88 40

Pipe stoves for heating blast.. 88 10

Pipes in ingots 35 10

** in steel. Prevention of. . . 85 11

Pit furnace 85 83

Pitting in steel 85 10

Plate column. Lunge 88 16

" mill 85 86

•• mill. Universal 85 87

** tower,Lunge,forcondens-

ing hydrochloric acid.. 80 6

Platinizing electrodes 80 51

Platinum black 87 46

Polarization, Electric 80 68

Ports to open-hearth furnace.. 83 13

Potassium chlorate 80 87

** chlorate. Analysis of 81 50
** chlorate. Apparatus
used in the manu-
facture of 80 88

** chlorate by Blumen-

berg electrolytic

process 80 86

** chlorate by electrol-
ysis 80 84

** chlorate by Gall-

and-Montlaur elec«

troly tic process... 80 85

xxii

INDEX

, Sec.
Potassium chlorate by Gibbs
electrolytic proc-
ess ao

" chlorate by the Cor-
b i n electrolytic

process 80

** chlorate, Determi-
nation of potas-
sium chloride in. . 81
** chlorate. Drying the

crystals of 80

** chlorate. Grinding

the crystals of 80

** chlorate. Lime used

in the manufac-

tureof 80

** chlorate. Process of

manufacture of . . . 80
** chlorate. Raw mate-
rials used in man-
ufacture of 80

** chlorate, Recrystal-

lizationof 80

** chloride, Determi-
nation of, in potas-
sium chlorate 81

Pots, Caustic 29

Potting 27

Pouring crucibles 84

Precipitation of arsenic in the

Freiberg process 88

Producer, Porter water-seal. . . 38
*' Fraser-Talbot me-
chanical 88

" gas 88

** gas. Calorific value

of 83

** gas, Composition of. 38
** gas, Fuel for making 88

'' • Operation of 83

" reactions 88

Siemens 88

Producers, Arrangement of.... 88

Water-seal 88

Protection of blast-furnace li-
nings 82

Pump, Kestner automatic 28

** Hontejus's with acid egg 28

Pumps, Acid 28

Purification of brine for Solvay

process 29

** of chamber acid... 28

^* of chamber acid

from arsenic 28

•* of hydrochloric

acid 80

Pajft Sec.

Purification of lye 20

Pyrite 82

87 Pyrites burners 27

*\ Copper and iron 27

Pyrrhotites, Copper-nickel 27

8ft

Q Sec.

Qualitative tests for arsenic in

50 hydrochloric acid 81

Quicklime, Analysis of 81

42 ** Determination of

calcium carbo-

42 natein 81

** Determination of

free calcium ox-

87 Ide in 81

** Determination of

88 insoluble matter

in 81

** Determination of

87 magnesia in 81

41 R Sec.
Reactions in blast furnace 82

** of the chamber proc-
60 ess for sulphuric

80 acid 26

25 " Producer 83

52 Reagents for Bunte burette ... 81

Recarbonization 84

4ft ** in the furnace 84

85 *' in the ladle... 84
Recarbonizers 84

86 Recarbonizing, Darby method

88 of 84

** in acid Besse-

42 mer process.. 84

41 ** with coal or

88 coke 84

40 Reciprocating type of furnace,

89 Spence 27

88 Recrystallization of potassium

48 chlorate 80

85 Red hematite 82

** hematite. Distribution of,

28 in United States 82

25 ** liquors. Analysis of 81

2ft ** liquors. Determination of
25 sodium sulphide, sul-
phite, thiosulphate, and
20 sulphate in 81

42 •* shortness 85

Reduction of ore in blast fur-

43 nace, 82

Refractories, Acid 85

7 " Basic 85

Paxre
50

5
27

Pa/re

40
5

Pasre
40

89
11
88
80
89
84

41
4

6
25

42
88
42

IKDBX

xxiii

Sec, Page

Refractories, Clay 85 88

Neutral 85 48

Silicious 85 41

Refractory materials 7& 87

** mixtures 85 45

Regenerative stoves for heat-
ing blast 82 21

Regenerator, Siemens 88 8

Reheating furnaces 86 83

•* of burner gas 27 52

Reich's test for sulphur dioxide

in burner gas 27 86

Resistance capacity 80 58

'* electrical, Measure-

mentof 80 47

Rhenania muffled type of fur-
nace 27 84

Roaster, Davis-Colby ore 82 18

" Grittinger ore 82 12

Roasting iron ores 82 0

*' Sources of loss of sul-
phur in 27 15

Rocksalt 20 8

Rolling furnace. Advantages of 88 17

'* furnace, Well man 88 15

" mills 85 84

Roof of open-hearth furnace. . . 38 7

S Sec. Fage

Salmirabile 20 85

Salt, Analysis of 81 17

cake 20 81

cake. Analysis of 81 10

cake. Apparatus and
method of manufacture

of 20 86

cake. Blind or muffle

roaster for 20 41

cake. Crude materials for 20 85

cake, Deacon's plus-pres-
sure for 20 41

cake. Determination of

. alumina in 81 10

cake. Determination of

ferric oxide in 81 10

cake. Determination of

free acid in 81 10

cake. Determination of

limein 81 20

cake. Determination of

magnesia in 81 20

cake. Determination of
matter Insoluble in

acids in 31 10

cake. Determination of

saltin 31 10

Salt

t»

tt
t(

ti
tt
tt

tt
t<
t(
tt

tt
tt
ti

Sec. Page

cake, Determinationof so-
dium sulphate in 81 20

cake. Mechanical furnaces

for 20 44

cake, Open roasters for . . 20 86

cake process. Analysis of
materials for 81 17

cake, Properties of 20 45

cake. Salt for making .... 30 35

cake. Sulphuric acid for

making 20 36

cake. Uses for ^9 45

cake. Yield of 20 45

Denaturated 20 0

Determination of, in am-
moniacal brine 31 7

Determination of, in
blackash 31 23

Determination of, in caus-
tic bottoms 81 85

Determination of, in caus-
tic liquor 31 34

Determination of, in
fishedsalts 81 34

Determination of, in lye
from extraction of black
ash 81 28

Determination of, in
mother liquor 31 14

Determination of, in salt
cake 81 10

Determination of sodium
chloridein 81 17

Determination of sulphur
trioxide in 31 18

Determination of water in 81 17

Electrolysis of 30 66

Electrolysis of, by Cast-
ner-Kellner process .... 80 78

Electrolysis of, by Green-
wood process 80 74

Electrolysis of, by Har-
grreave-and-Bird proc-
ess 80 77

Electrolysis of, by proc-
esses using a mercury
cathode , 80 78

Electrolysis of, by proc-
esses using diaphragrms 30 74

Electrolysis of, by the Le

Sueur process 80 75

Electrolysis of, with dis-
solved electrolyte 80 78

Electrolysis of, with fused
electrolyte 30 60

for making salt cake 20 85

XXIV

INDEX

Sec. Page

Salt from brine 29 4

** from Bea- water 29 1

** Occurrence of 89 1

** Rock 29 8

Sand for making crucibles. .... 84 48
Saniter process for desulphuri-
zation in baste open-hearth

process 88 88

Saturator for makmg ammo-

niacal brine 89 80

Scaffolds m blast furnace 80 64

Sea-water, Salt from 89 1

Segregation m steel castings. . 85 5
Selenium, Determination of« in

hydrochloric acid 81 40

Settling pans tor manufacture

of potassium chlorate 80 40

Shank's lixiviation system of

Wackash 89 67

Shunt circuit....'. 80 58

Side blowing 84 18

Siemens producer 33 33

" regenerator 88 8

valve 88 80

Silica, Determination of, in lye
from extraction of

blackash 81 84

** Determination of. In

soda ash — 81 16

Silicious refractory materials.. 35 41

Silicon, Effect of, in steel ^ 14

in iron 32 . 57

** Loss of, in recarf>oni-

zing 81 40

** Oxidation of, in basic

Bessemer process.... 84 87

Silicospiegel 84 84

Skew back 83 8

Slab 85 84

SlabbingmiU 85 87

Slag, Itosic open-hearth 33 74

*' Composition of blast-fur-
nace 82 45

** in acid open-hearth proc-
ess 33 53

Slags, Blast-furnace &} 45

" Fusibility of blast-fur-
nace 82 46

•♦ Handlingof 82 47

Slaked lime. Analysis of 81 48

** lime. Determination of

carbon dioxide in 81 48

" lime. Determination of

water in 81 48

Slipping in blast furnace 82 65

Soakmgpit 85 38

Sec.
Soda, ammonia. Analysis of. . . 81
ammonia. Properties of. 89

Artificial 89

ash, Analysis oC 81

ash, Analysis of 81

ash. Determination of

caustic soda in 81

aah, Determination of fer-
ric oxide and alumina in
ash. Determination of
magnesium carbonate

ash. Determination of

silica in 81

ash, Determination of so-
dium bicarbonate in...
ash, Determination of so-
dium carbotaate in

ash, Determination of so-
dium Carbonate in

ash. Determination of so-
dium chloride in 81

ash. Determination of so-
dium sulphate in 81

ash. Determination of so-
dium sulphide in 31

ash, Determination of to-
tal alkali in 81

ash. Finished 83

ash. Grinding of 89

ash, Methods of stating

strength of 89

ash used in the manufac-
ture of sodium hydrate 29
by the Le Blanc process. 89

caustic. Analysis of 81

crystal , Anal y sis of 81

crjTstals 89

crystals, Calcination of.. 80

Natural 89

process. Cryolite 89

Sodium aluminate from cryo-
lite soda process... . 89

*' bicarbonate 89

** bicarbonate. Analysis

of 81

** bicarbonate, Calcina-
tion of 89

bicarbonate. Determi-
nation of, in bicarbo-
nate from filters 81

bicarbonate, Determi-
nation of, in carbo-
nated lye 81

bicarbonate. Determi-
nation of, in soda ash 81

Page
1

31
11
16
27

81 16

t»

!•

t»

88
64
68

75
46
86
89

82
86

85
16

INDEX

XXV

Sec,
Sodium bicarbonate, Dry proc-
ess tor purification of t9
bicarbonate, Wet proc-
ess for purification of 20

** carbonate 29

** carbonate, Caustici*

zing SO

*' carbonate, Crystals of 99

'' carbonate. Determina-
tion of, in black ash.. 81

'^ carbonate, Determina-
tion of, in caustic bot-
toms 81

** carbonate, Determina-
tion of, in caustic
liquor 81

^* carbonate. Determina-
tion of, in lye from
extraction of black
ash 81

** carbonate, Determina-
tion of, in soda ash . . "SI

'* carbonate, Determina-
tion of, in soda ash . . 81

** carbonate. Uses of.... 29

** chlorate 30

" chloride 89

*' chloride. Determina-
tion of, in brine 81

*• chloride. Determina-
tion of, in salt 81

** chloride. Determina-
tion of, in soda ash . . 81

** compounds, total. De-
termination of, in
tank waste 81

'* ferrocyanide. Deter-
mination of, in lye
from extraction of
black ash 81

" hydrate 29

'* hydrate. Crude mate-
rials used in manu-
facture of 29

** hydrate. Details of
process of manufac-
ture of 29

*" hydrate. Evaporation

of 29

** hydrate. Lime used in

the manufacture of.. 29

" hydrate, Loe wig's
process for manufac-
ture of 29

** hydrate, Removal of

sulphur from 29

Page

84
9

76
08

38
85

24
74

76
78
75

88
81

*•

k»

84
25
20
17
47
22

Sec. Page
Sodium hydrate. Soda ash used

in the manufacture of 29 75

hydrate. Uses of 29 83

*' sulphate 29 ai

*' sulphate. Determina-
tion of, in black ash. 81 23
** sulphate. Determina-
tion of, in fished salts 31 84
sulphate. Determina-
tion of, in lye from
extraction of black

ash 81

sulphate. Determina-
tion of, in red liquors 31
sulphate, Determina-
tion of, In salt cake.. 31
sulphate, Determma-

tion of, in soda ash . . 31
sulphate for Le Blanc

soda process 89

sulphide. Determina-
tion of, in black ash. 31
sulphide. Determina-
tion of, in lye from
extraction of black

ash 81 28

sulphide. Determina-
tion of, in red liquors 81 25
sulphide. Determina-
tion of, in soda ash.. 31 2R

thiosulphate 29 73

thiosnlphate. Determi-
nation of, in red

liquors 31 25

Softball 29 51

Solutions for resistance capac-
ity 30 M

Solvay process 29 11

process. Ammonia lost

in 29 31

process. Ammonia re-
covery in 29 28

process. Ammonia used

in , 29 14

process, Aramoniacal

brine of 29 20

process. Brine used in.. 29 14
process. Carbon diox-
ide for 29 15

process, Carbonating

ammoniacal brine for 29 21
process, Carbonating

tower for 29 22

process, Coal and coke

used in 29 15

process, Details of. ... 29 15

XXVI

INDEX

Sec. Page
Solvay process, Distiller liquor

in » 80

*' process, Filters for 89 94

*^ process, Limestone

used in M 18

** process, Ptirification of

brine for 29 20

^* process, Raw materials

used in 29 18

** process. Washing of

carbon droxide for... 29 18

Sows 82 40

Spathic ore... 82 6

Specific gravity. Determina-
tion of, of lye from
extraction of black

ash 31 23

*• gravity of milk of
lime. Determination

of 81 14

** gravity of ammonia
liquor, Determina-
tion of 81 6

** gravity of brine. De-
termination of 81 1

'* gravity of caustic
liquor, Determina-
tion of 81 88

** gravity of sulphuric
acid. Determination

of 27 6

** gravity scale, Euro-
pean, Baum6 27 0

** gravity scale, Twad-

dell 27 6

** gravity scale, United

States Baumd 27 6

Spence reciprocating type of

furnace 27 81

Spiegeleisen 82 XA

" M 84

Stahl method for removing ar-
senic from chamber acid 28 IT

Steam, Admission of, to the

lead chambers....? 28 33

Steel, Air-quenched 84 fiO

" castings 85 1

'* castings, Blowholes in . . 85 9
** castings, Composition of 85 3
** castings, Segregation in 85 5
*• castings, Solidity of.. . 85 2
** Chemical examination of 85 16
** crucible. Composition of 84 64

" Defects in 35 6

'* Definition of 33 1

*' Effect of carbon in 85 12

Sec.
Steel, Effect of copper, nickel,

and aluminum in 85

** Effect of manganese in. . 85
** Effect of oxides or oxy-
gen in 85

** Effect of phosphorus in. 85

" Effect of silicon in 85

'' Effect of sulphur in 85

" Effects of work and heat

on .'.. 86

'* Effects of usual elements

present in 85

*' Had field's manganese... 84
^* History of manufacture

of 83

** ladle 83

'• ladle 84

** making, Recent progress

in 35

** Manganese 84

** Microscopical examina-
tion of 85

" mills 85

** Nickel 84

** Open-hearth process for 88

** Physical testing of 85

** Processes of manufac-
ture of 83

** Properties of, deter-
mined in testing 85

*■* Relation of chemical
composition to strength

of 86

*■*" Results in physical test-
ing of 36

" Superiority of crucible.. 84

** Test pieces of 35

*• Tungsten 84

Steels, Alloy 84

*^ Aluminum, copper, and

titanium 34

Still for decomposition of hy-
drochloric acid by man-
ganese dioxide 30

" for recovery of ammonia

in Solvay process 29

'* liquor, Determination of

free acid in 81

** liquors from the decompo-
sition of hydrochloric acid
by manganese dioxide. .. 80

Stopper head 88

Stoves for heating blast 82

'' Pipe, for heating blast. 82
** Regenerative, for heat-
ing blast 82

Page

14
12

14
18
14
13

II
59

2
27

15
22
59
6
16

25
63
18
5«
5ft

12
89

15
27
19
19

INDEX

XXVll

Sec. Page
Strength of solutions weaker
than the monohydrate, Com-
mercial methods for deter*

mining the 87 5

Sulphides, metallic. Prepara-
tion of 27 11

Sulphur, Available 87 15

** ' available. Determi-
nation of, in tank

waste 81 89

** Available, in burner

gas 87 14

" Combustion of.., 27 12

** compounds, oxidiza-
ble, Determination
of, in fished salts.... 81 84
** dioxide or burner gas.

Production of 27 21

'' dioxide, Reich's test

for in burner gas. . . 27 86

** Effect of, in steel 85 18

** Grading of crude 27 0

" iniron 82 69

** in tank waste, RecoV'
ery of, by Chance-

Claus process 29 69

** Oxidation of, in basic

Bessemer process... 34 28

** recovered 27 8

** Removal of, from Le

Blanc caustic soda. . 29 81
** Sources of loss of, in

roasting 27 15

" Thermochemistry of

the combustion of. . 27 12
** total, Determination
of, in lye from ex-
traction of black ash 81 24
" trioxide. Determina-
tion of. in salt 81 18

** trioxide. Hydrates

and solutions of 27 1

** trioxide in brine, De-
termination of 81 8

** trioxide, Monohy-
drate of 27 8

" trioxide. Nomencla-
ture of solutions and

hydrates of 27 3

Sulphureted hydrogen 27 8

** hydrogen gener-
ator, Freiberg. 28 44
Sulphuric acid, Catalytic or
contact process for
the manufacture
of 27 43

Sec. Page
Sulphuric acid, Chamber proc-
ess for 28 1

** acid. Concentration
and distillation of,
starting with the
Glover tower 28 56

" acid. Concentration
of, by the Kessler
process 28 56

** acid, Concentration
of dilute solutions
of 28 48

" acid, Concentration
of, in glass beak-
ers or dishes 28 58

** acid. Concentration
of, in glass retorts
orstills 28 52

** acid. Concentration

of, in iron 28 51

** acid. Concentration

of, in lead pans... 28 48

'* acid. Concentration

of, in platinum... 28 49

** acid. Conditions of,

in the chambers., 28 80

** acid. Conditions of,
i n the Glover
tower — .... 28 29

** acid. Contact mass
or material used in
the manufacture
of, by the contact
process.. 27 45

** acid, Control of
chamber process
for 28 88

'* acid, Definition of.. 27 8

*' acid. Determination
of, in hydrochloric
acid 81 89

'^ acid. Determination
of specific gravity
or density of 27 5

** acid. Diagram of

manufacture of... 28 60

** acid for making salt

cake 29 86

'* acid,Nordhausenor

fuming 27 4

** acid. Operation of
chamber process
for 26 28

'* acid. Preparation of

raw material for
manufacture of... 27 9

xxvni

INDEX

Sec,
Sulphuric acid, Principlesgov-
erning: the manu-
facture of 27

* ' acid, Raw materials

used in the manu>

fact u re of 27

*^ acid. Reactions of

the chamber proc-

ess for 28

'' acid, SUrting the

chamber process

for 28

** hydrate. Yield and

method of calcu-
lating yield of.... 27
** monohydrate.

Lunge freezing
process for the

production of 28

Sulphurous acid. Determina-
tion of, in hydrochloric acid. 31

Surface condensers 28

** heat evaporation of

tank liquor 29

T Sec.

Talbot continuous open-hearth

process 86

Tank liquor. Evaporation of... 29
*' liquor, Evaporation of, in

mechanical pans 29

** liquor, Evaporation of, in

pans with heat below . . 29
** liquor. Evaporation of,

with surface heat 29

** waste, Analysis of 81

** waste, Determination of

alkalina compounds in. 81
** waste, Determination of

available sulphur in. . .. 81
** waste, Determination of
total sodium com-
pounds in 81

** waste in Le Blanc process 29

Teeming 83

" crucibles 84

Temperature, Allowance for,
in determining
the Baum6
gravity of sul-
phuric acid .... 27
** in converter of

acid Bessemer

process 34

Tensile-strength test of steel.. 85
Test pieces of steel 85

Pa^e

89
16

Pa^e

60
26

27
67
27
52

17
20

Sec.

Testing burner gas 27

*' machine 36

Thenardite 29

Thermochemistry of the com-
bustion of sulphur 27

Thomas-Gilchrist process 34

Titanium in iron , 9Si

Total alkali, Determination of,

in bicarbonate from filters... 31
Tower, Carbonating, for Sol-

vay process 29

*' Gay-Lussac 28

" Glover's 28

Towers,Gay-Lus8ac,Numberof 28

Tropenas process 84

Tungsten steel 84

Tuyeres, Blast-furnace 82

** for cupolas 84

** of acid Bessemer con-
verter 34

*• of basic converter.... 34
** To detect leak ing . . . . 80
** taking blast irregu-
larly 30

Twaddell specific-gravity scale 27

U Sec.

Units of electrical measurement 30

Universal mill 85

»• plate mill 85

V Sec.
Vacuum pan process for evapo-
rating brine 29

Valuation of bleach 30

Valve,Forter 88

" Siemens 88

Valves, Ctit-off gas 83

Varil 29

Vessel patching 85

Vitriol, Nitrous 28

Voltmeter 80

W Sec.

Washing gases from ammonia

saturator 29

" iron ores 82

Waste from ammonia stills.

Analysis of 81

*' gas from Claus kiln,

Analysis of 81

*' gases from hydrochlo-
ric - acid absorption,

Analysis of. 81

Water^Determination of, in salt 81
** Determination of, in

slaked lime 81

Pajze
' 86

12
21
GO

8
3?
81
50
29

14
28
62

63
0

Pa/re
46
37
87

Pa^e

6
33
21
20
22
10
45

5
58

PajS^^

14
29

86
17

INDEX

XXIX

Sec.

Water gas 88

"'' seal producer 33

'* seal producer, Forter... 83
Weldon and Deacon processes
for chlorine, Com-
parison of 80

mud 80

" mud, Analysis of 81

" mud. Determination

of manganese in 81

" mud. Determination

of total base fn 81

•* process for chlorine... 80
•* process for the recov-
ery of mansranese
from still liquors .... 80

Page
84
36
85

27
10
44

Sec. Page

Wellman charging machine... 88 25

" rolling furnace 88 16

" rolling furnace,

Porehearthof 88 16

Wheatstone bridge 80 47

Whitwell stove -. 82 28

Woots steel 88 8

Y Sec. Page

Yaryan evaporator 29 78

Yield, and method of calcula-
ting yield of sulphuric hy-
drate^ 27* 17

2 Sec. Page

ZIncblende 27 8

168 >B/w

7009

6^071'^li275

b89071«1«7**

LlRo.

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Sulphuric acid ; Alkalies and hydrochloric acid ; Manufacture of iron ; Manufacture of steel"

See other formats